https://blog.schoergendorfer.com/post/cloud-and-on-premises-deployment-of-uzbek-legal-rag-via-targeted-retriever-fineSecFeedde2026-09-01T18:32:39.981282+00:00Cloud and On-Premises Deployment of Uzbek Legal RAG via Targeted Retriever Fine-Tuninghttps://blog.schoergendorfer.com/post/modality-fault-lines-structural-corruptions-reveal-fragile-omni-modal-reasoningSecFeedde2026-09-01T18:32:39.958976+00:00Modality Fault Lines: Structural Corruptions Reveal Fragile Omni-Modal Reasoninghttps://blog.schoergendorfer.com/post/large-language-models-systematically-favor-popular-options-evidence-and-mitigatiSecFeedde2026-09-01T18:32:39.933420+00:00Large Language Models Systematically Favor Popular Options: Evidence and Mitigation Across MCQshttps://blog.schoergendorfer.com/post/when-patients-cut-in-extending-clinical-conversational-ai-safety-to-interruptionSecFeedde2026-09-01T18:32:39.912478+00:00When Patients Cut In: Extending Clinical Conversational AI Safety to Interruptionshttps://blog.schoergendorfer.com/post/anchoring-speech-with-semantics-a-multimodal-adapter-mechanism-for-automatic-speSecFeedde2026-09-01T18:32:39.892982+00:00Anchoring Speech with Semantics: A Multimodal Adapter Mechanism for Automatic Speech Recognition in Low-Resource Languageshttps://blog.schoergendorfer.com/post/attribute-based-activation-steering-of-llms-for-group-specific-explanation-generSecFeedde2026-09-01T18:32:39.870946+00:00Attribute-Based Activation Steering of LLMs for Group-Specific Explanation Generationhttps://blog.schoergendorfer.com/post/toward-cultural-alignment-human-centered-evaluation-of-multimodal-ai-stories-acrSecFeedde2026-09-01T18:32:39.849845+00:00Toward Cultural Alignment: Human-Centered Evaluation of Multimodal AI Stories Across Five African Communitieshttps://blog.schoergendorfer.com/post/toward-a-cross-lingual-romanization-ecosystem-for-sinitic-languages-a-paired-manSecFeedde2026-09-01T18:32:39.829516+00:00Toward a Cross-Lingual Romanization Ecosystem for Sinitic Languages: A Paired Mandarin-Cantonese Case Studyhttps://blog.schoergendorfer.com/post/quantifying-error-tolerance-in-synthetic-data-an-atomic-level-operand-vs-operatoSecFeedde2026-09-01T18:32:39.810631+00:00Quantifying Error Tolerance in Synthetic Data: An Atomic-level Operand vs. Operator Perturbation Studyhttps://blog.schoergendorfer.com/post/ai-historian-helping-historians-organize-and-verify-person-centred-temporal-clueSecFeedde2026-09-01T18:32:39.791836+00:00AI Historian: Helping historians organize and verify person-centred temporal clues from dispersed historical narrativeshttps://blog.schoergendorfer.com/post/recognition-refusal-misalignment-in-llms-why-models-answer-structurally-unanswerSecFeedde2026-09-01T18:32:39.769511+00:00Recognition-Refusal Misalignment in LLMs: Why Models Answer Structurally Unanswerable Questionshttps://blog.schoergendorfer.com/post/routesparse-input-conditional-pattern-routing-for-budgeted-long-context-prefilliSecFeedde2026-09-01T18:32:39.748438+00:00RouteSparse: Input-Conditional Pattern Routing for Budgeted Long-Context Prefillinghttps://blog.schoergendorfer.com/post/detecting-and-guiding-llm-generated-korean-poetry-with-interpretable-form-levelSecFeedde2026-09-01T18:32:39.730167+00:00Detecting and Guiding LLM-Generated Korean Poetry with Interpretable Form-level Featureshttps://blog.schoergendorfer.com/post/cova-sft-a-large-scale-dataset-for-chain-of-visual-abstractionsSecFeedde2026-09-01T18:32:39.709915+00:00CoVA-SFT: A Large-Scale Dataset for Chain of Visual Abstractionshttps://blog.schoergendorfer.com/post/vocalaffectbench-evaluating-vocal-emotion-recognition-in-ai-audio-modelsSecFeedde2026-09-01T18:32:39.684454+00:00VocalAffectBench: Evaluating Vocal Emotion Recognition in AI Audio Modelshttps://blog.schoergendorfer.com/post/voicecodebench-evaluating-exact-structured-token-recovery-in-automatic-speech-reSecFeedde2026-09-01T18:32:39.626579+00:00VoiceCodeBench: Evaluating Exact Structured-Token Recovery in Automatic Speech Recognitionhttps://blog.schoergendorfer.com/post/evaluating-multilingual-sentence-embeddings-for-translation-error-detectionan-enSecFeedde2026-09-01T18:32:39.586203+00:00Evaluating Multilingual Sentence Embeddings for Translation Error Detection:An English--Greek Contrastive Studyhttps://blog.schoergendorfer.com/post/reva-a-region-aware-visual-assistant-for-visually-grounded-question-answeringSecFeedde2026-09-01T18:32:39.559018+00:00ReVA: A Region-Aware Visual Assistant for Visually Grounded Question Answeringhttps://blog.schoergendorfer.com/post/forecasting-weather-driven-price-dynamics-across-sri-lankan-tea-market-catalogue-2SecFeedde2026-09-01T18:18:05.978994+00:00Forecasting Weather-Driven Price Dynamics Across Sri Lankan Tea Market Catalogueshttps://blog.schoergendorfer.com/post/satdl-jointly-optimizing-data-redistribution-and-training-for-satellite-based-di-2SecFeedde2026-09-01T18:18:05.959362+00:00SatDL: Jointly Optimizing Data Redistribution and Training for Satellite-Based Distributed Learninghttps://blog.schoergendorfer.com/post/gta-rag-graph-trajectory-augmented-reinforcement-learning-for-multi-turn-retriev-2SecFeedde2026-09-01T18:18:05.934343+00:00GTA-RAG: Graph-Trajectory-Augmented Reinforcement Learning for Multi-Turn Retrieval-Augmented Reasoninghttps://blog.schoergendorfer.com/post/a-comprehensive-review-of-large-language-models-for-nanophotonics-from-surrogate-2SecFeedde2026-09-01T18:18:05.913155+00:00A Comprehensive Review of Large Language Models for Nanophotonics: From Surrogate Modeling to Autonomous Designhttps://blog.schoergendorfer.com/post/a-deterministic-constant-competitive-algorithm-for-dynamic-mixture-of-experts-seSecFeedde2026-09-01T18:18:05.886888+00:00A Deterministic Constant-Competitive Algorithm for Dynamic Mixture-of-Experts Servinghttps://blog.schoergendorfer.com/post/mitre-sage-a-multi-agent-cybersecurity-question-answering-model-3SecFeedde2026-09-01T18:18:05.867316+00:00MITRE-SAGE: A Multi-Agent Cybersecurity Question-Answering Modelhttps://blog.schoergendorfer.com/post/sdf-aware-weighting-adaptive-eikonal-regularisation-for-three-dimensional-levelSecFeedde2026-09-01T18:18:05.847132+00:00SDF-Aware Weighting: Adaptive Eikonal Regularisation for Three-Dimensional Level-Set Physics-Informed Neural Networkshttps://blog.schoergendorfer.com/post/fast-trainable-multilinear-bases-for-image-compression-3SecFeedde2026-09-01T18:18:05.814556+00:00Fast Trainable Multilinear Bases for Image Compressionhttps://blog.schoergendorfer.com/post/the-value-of-depth-in-message-passing-on-sparse-graphs-a-kesten-stigum-dichotomy-2SecFeedde2026-09-01T18:18:05.796022+00:00The Value of Depth in Message Passing on Sparse Graphs: A Kesten-Stigum Dichotomyhttps://blog.schoergendorfer.com/post/regionfm-interpretable-region-based-brain-mri-classification-using-foundation-mo-2SecFeedde2026-09-01T18:18:05.777757+00:00RegionFM: Interpretable Region-Based Brain MRI Classification Using Foundation Model Embeddingshttps://blog.schoergendorfer.com/post/posterior-variance-is-a-constraint-map-not-an-error-map-closed-form-uncertainty-2SecFeedde2026-09-01T18:18:05.760608+00:00Posterior Variance Is a Constraint Map, Not an Error Map: Closed-Form Uncertainty for Radiative Gaussian Splatting in Sparse-View CThttps://blog.schoergendorfer.com/post/runtime-safety-filtering-for-learned-small-uas-separation-policies-under-gnss-de-2SecFeedde2026-09-01T18:18:05.739709+00:00Runtime Safety Filtering for Learned Small UAS Separation Policies under GNSS Degradationhttps://blog.schoergendorfer.com/post/fracevent-event-camera-simulation-via-fractional-relaxation-pixel-dynamicsSecFeedde2026-09-01T18:18:05.695320+00:00FracEvent: Event-Camera Simulation via Fractional-Relaxation Pixel Dynamicshttps://blog.schoergendorfer.com/post/dont-read-everything-a-curvature-conditioned-query-for-linear-attentionSecFeedde2026-09-01T18:18:05.674548+00:00Don't Read Everything: A Curvature-Conditioned Query for Linear Attentionhttps://blog.schoergendorfer.com/post/privacy-enhanced-zero-order-federated-learning-via-xmk-ckks-over-wireless-channeSecFeedde2026-09-01T18:18:05.654261+00:00Privacy-Enhanced Zero-Order Federated Learning via xMK-CKKS over Wireless Channelshttps://blog.schoergendorfer.com/post/correcting-test-set-contamination-by-spiking-the-training-dataSecFeedde2026-09-01T18:18:05.616924+00:00Correcting test set contamination by spiking the training datahttps://blog.schoergendorfer.com/post/physics-guided-concentration-inference-from-resistance-transients-in-a-mixed-pha-2SecFeedde2026-09-01T18:02:54.527577+00:00Physics-Guided Concentration Inference from Resistance Transients in a Mixed-Phase SnO-SnO$_2$ Carbon Monoxide Sensor with p-n Switchinghttps://blog.schoergendorfer.com/post/operator-guided-model-reduction-for-generative-sampling-in-lattice-field-theorySecFeedde2026-09-01T18:02:54.508681+00:00Operator-Guided Model Reduction for Generative Sampling in Lattice Field Theoryhttps://blog.schoergendorfer.com/post/learning-the-channel-gain-from-anywhere-to-anywhere-via-cross-environment-transfSecFeedde2026-09-01T18:02:54.489831+00:00Learning the Channel Gain from Anywhere to Anywhere via Cross-environment Transformer Estimatorshttps://blog.schoergendorfer.com/post/large-language-model-enabled-automated-data-extraction-for-concrete-materials-inSecFeedde2026-09-01T18:02:54.469578+00:00Large language model-enabled automated data extraction for concrete materials informaticshttps://blog.schoergendorfer.com/post/performance-manipulation-labor-market-implications-in-ai-assisted-eraSecFeedde2026-09-01T18:02:54.451931+00:00Performance Manipulation: Labor Market Implications in AI-assisted Erahttps://blog.schoergendorfer.com/post/verix-anon-a-multi-layered-framework-for-mathematically-verifiable-outsourced-ta-2SecFeedde2026-09-01T18:02:54.433446+00:00VeriX-Anon: A Multi-Layered Framework for Mathematically Verifiable Outsourced Target-Driven Data Anonymizationhttps://blog.schoergendorfer.com/post/synmulti-synthetic-to-real-learning-for-multimodal-video-understandingSecFeedde2026-09-01T18:02:54.413314+00:00SynMulti: Synthetic-to-Real Learning for Multimodal Video Understandinghttps://blog.schoergendorfer.com/post/operator-learning-for-predicting-bulk-wave-parameters-of-spectral-wave-modelsSecFeedde2026-09-01T18:02:54.396636+00:00Operator Learning for Predicting Bulk Wave Parameters of Spectral Wave Modelshttps://blog.schoergendorfer.com/post/model-selection-and-parameter-estimation-for-multidimensional-gaussian-mixture-mSecFeedde2026-09-01T18:02:54.361784+00:00Model Selection and Parameter Estimation for Multidimensional Gaussian Mixture Models with a Common Covariance Matrixhttps://blog.schoergendorfer.com/post/nanovdr-distilling-a-2b-vision-language-retriever-into-a-70m-text-only-encoder-fSecFeedde2026-09-01T18:02:54.342478+00:00NanoVDR: Distilling a 2B Vision-Language Retriever into a 70M Text-Only Encoder for Visual Document Retrievalhttps://blog.schoergendorfer.com/post/prediction-powered-conditional-inferenceSecFeedde2026-09-01T18:02:54.323283+00:00Prediction-Powered Conditional Inferencehttps://blog.schoergendorfer.com/post/ai-generated-measurements-for-identification-and-inference-with-missing-data-a-wSecFeedde2026-09-01T18:02:54.306408+00:00AI-Generated Measurements for Identification and Inference with Missing Data: A Weak Shadow Variable Approachhttps://blog.schoergendorfer.com/post/edge-local-and-qubit-efficient-quantum-graph-learning-for-the-nisq-era-3SecFeedde2026-09-01T18:02:54.290570+00:00Edge-Local and Qubit-Efficient Quantum Graph Learning for the NISQ Erahttps://blog.schoergendorfer.com/post/efficient-reduction-of-stellar-contamination-and-noise-in-planetary-transmissionSecFeedde2026-09-01T18:02:54.270332+00:00Efficient reduction of stellar contamination and noise in planetary transmission spectra using neural networkshttps://blog.schoergendorfer.com/post/diverse-via-bounded-agreement-geometric-regularization-for-multimodal-fusionSecFeedde2026-09-01T18:02:54.252858+00:00Diverse via bounded Agreement: Geometric Regularization for Multimodal Fusionhttps://blog.schoergendorfer.com/post/unknown-unknowns-do-hidden-intentions-in-llms-evade-detectionSecFeedde2026-09-01T18:02:54.236092+00:00Unknown Unknowns: Do Hidden Intentions in LLMs Evade Detection?https://blog.schoergendorfer.com/post/tracing-the-latent-threads-a-mechanistic-study-of-how-llms-represent-and-operatiSecFeedde2026-09-01T18:02:54.185149+00:00Tracing the Latent Threads: A Mechanistic Study of How LLMs Represent and Operationalize Race and Ethnicity Cueshttps://blog.schoergendorfer.com/post/x-coder-advancing-competitive-programming-with-synthetic-tasks-solutions-and-tesSecFeedde2026-09-01T17:47:50.358384+00:00X-Coder: Advancing Competitive Programming with Synthetic Tasks, Solutions, and Testshttps://blog.schoergendorfer.com/post/soft-fitted-q-iteration-without-bellman-completeness-occupancy-reweighting-and-tSecFeedde2026-09-01T17:47:50.337268+00:00Soft Fitted Q-Iteration without Bellman Completeness: Occupancy Reweighting and Temperature Annealinghttps://blog.schoergendorfer.com/post/fitted-q-evaluation-without-bellman-completeness-via-occupancy-weightingSecFeedde2026-09-01T17:47:50.311595+00:00Fitted Q-Evaluation without Bellman Completeness via Occupancy Weightinghttps://blog.schoergendorfer.com/post/spider-spatially-informed-dense-embedding-retrieval-for-software-issue-localizatSecFeedde2026-09-01T17:47:50.276301+00:00SpIDER: Spatially Informed Dense Embedding Retrieval for Software Issue Localizationhttps://blog.schoergendorfer.com/post/tpso-training-free-diverse-image-generation-via-semantic-prompt-embedding-optimiSecFeedde2026-09-01T17:47:50.257163+00:00TPSO: Training-Free Diverse Image Generation via Semantic Prompt Embedding Optimizationhttps://blog.schoergendorfer.com/post/dcc-data-centric-compilation-of-machine-learning-kernels-for-processing-in-memor-2SecFeedde2026-09-01T17:47:50.236632+00:00DCC: Data-Centric Compilation of Machine Learning Kernels for Processing-In-Memory Architectureshttps://blog.schoergendorfer.com/post/perk-long-context-reasoning-as-test-time-learningSecFeedde2026-09-01T17:47:50.215635+00:00PERK: Long-Context Reasoning as Test-Time Learninghttps://blog.schoergendorfer.com/post/mirror-descent-linearized-augmented-lagrangian-methods-for-nonconvex-constrainedSecFeedde2026-09-01T17:47:50.181581+00:00Mirror Descent Linearized Augmented Lagrangian Methods for Nonconvex Constrained Stochastic Zeroth-Order Optimizationhttps://blog.schoergendorfer.com/post/optimal-estimation-of-watermark-proportions-in-hybrid-ai-human-textsSecFeedde2026-09-01T17:47:50.159698+00:00Optimal Estimation of Watermark Proportions in Hybrid AI-Human Textshttps://blog.schoergendorfer.com/post/trust-under-siege-label-spoofing-attacks-against-machine-learning-for-android-maSecFeedde2026-09-01T17:47:50.138046+00:00Trust Under Siege: Label Spoofing Attacks against Machine Learning for Android Malware Detectionhttps://blog.schoergendorfer.com/post/perforated-backpropagation-a-neuroscience-inspired-extension-to-artificial-neuraSecFeedde2026-09-01T17:47:50.110831+00:00Perforated Backpropagation: A Neuroscience Inspired Extension to Artificial Neural Networkshttps://blog.schoergendorfer.com/post/classification-drives-geographic-bias-in-street-scene-segmentationSecFeedde2026-09-01T17:47:50.084045+00:00Classification Drives Geographic Bias in Street Scene Segmentationhttps://blog.schoergendorfer.com/post/autoencoders-in-function-spaceSecFeedde2026-09-01T17:47:50.064273+00:00Autoencoders in Function Spacehttps://blog.schoergendorfer.com/post/learning-diverse-attacks-on-large-language-models-for-robust-red-teaming-and-safSecFeedde2026-09-01T17:47:50.038379+00:00Learning diverse attacks on large language models for robust red-teaming and safety tuninghttps://blog.schoergendorfer.com/post/model-selection-and-parameter-estimation-of-one-dimensional-gaussian-mixture-modSecFeedde2026-09-01T17:47:50.018746+00:00Model Selection and Parameter Estimation of One-Dimensional Gaussian Mixture Modelshttps://blog.schoergendorfer.com/post/phylogfn-phylogenetic-inference-with-generative-flow-networksSecFeedde2026-09-01T17:47:49.980240+00:00PhyloGFN: Phylogenetic inference with generative flow networkshttps://blog.schoergendorfer.com/post/deep-graph-kernel-point-processes-over-networksSecFeedde2026-09-01T17:47:49.959382+00:00Deep graph kernel point processes over networkshttps://blog.schoergendorfer.com/post/beyond-parallel-blindness-information-floors-and-model-gaps-in-block-drafting-2SecFeedde2026-09-01T17:32:44.490084+00:00Beyond Parallel Blindness: Information Floors and Model Gaps in Block Draftinghttps://blog.schoergendorfer.com/post/quantumboostnet-hybrid-classical-quantum-cardiac-view-identificationSecFeedde2026-09-01T17:32:44.466127+00:00QuantumBoostNet: Hybrid Classical-Quantum Cardiac View Identificationhttps://blog.schoergendorfer.com/post/a-unified-framework-for-fair-and-personalized-decentralized-learning-under-commu-2SecFeedde2026-09-01T17:32:44.448332+00:00A Unified Framework for Fair and Personalized Decentralized Learning under Communication Constraintshttps://blog.schoergendorfer.com/post/frequency-aware-forecasting-for-short-term-typhoon-gust-prediction-2SecFeedde2026-09-01T17:32:44.429250+00:00Frequency-aware forecasting for short-term typhoon gust predictionhttps://blog.schoergendorfer.com/post/every-layer-counts-an-exponential-l_2-depth-hierarchy-for-relu-networks-2SecFeedde2026-09-01T17:32:44.407278+00:00Every Layer Counts: An Exponential $L_2$ Depth Hierarchy for ReLU Networkshttps://blog.schoergendorfer.com/post/catchbench-when-can-an-agent-failure-be-caughtSecFeedde2026-09-01T17:32:44.388190+00:00CatchBench: When Can an Agent Failure Be Caught?https://blog.schoergendorfer.com/post/in-cell-learning-language-models-that-update-their-own-weights-in-sequence-withoSecFeedde2026-09-01T17:32:44.352554+00:00In-Cell Learning: Language Models That Update Their Own Weights in Sequence Without Changing the File They Shiphttps://blog.schoergendorfer.com/post/machine-learning-and-arima-model-averaging-for-adaptive-public-health-forecastin-2SecFeedde2026-09-01T17:32:44.332189+00:00Machine Learning and ARIMA Model Averaging for Adaptive Public Health Forecasting: Comparative Evaluation and an Ontario COVID-19 Case Studyhttps://blog.schoergendorfer.com/post/large-discovery-models-empirically-grounded-model-based-open-ended-search-2SecFeedde2026-09-01T17:32:44.310393+00:00Large Discovery Models: Empirically-grounded Model-Based Open-Ended Searchhttps://blog.schoergendorfer.com/post/momentum-as-residual-driven-multiplier-correction-for-deep-learning-optimization-2SecFeedde2026-09-01T17:32:44.287047+00:00Momentum as Residual-Driven Multiplier Correction for Deep Learning Optimizationhttps://blog.schoergendorfer.com/post/scaling-automatic-research-agents-via-world-models-2SecFeedde2026-09-01T17:32:44.269139+00:00Scaling Automatic Research Agents via World Modelshttps://blog.schoergendorfer.com/post/when-do-task-vectors-interfere-mapping-the-validity-boundaries-of-weight-space-c-2SecFeedde2026-09-01T17:32:44.248712+00:00When Do Task Vectors Interfere? Mapping the Validity Boundaries of Weight-Space Compositionhttps://blog.schoergendorfer.com/post/task-to-model-optimization-for-enterprise-llm-coding-assistants-a-data-driven-fr-2SecFeedde2026-09-01T17:32:44.230173+00:00Task-to-Model Optimization for Enterprise LLM Coding Assistants: A Data-Driven Framework for Cost-Optimal Routinghttps://blog.schoergendorfer.com/post/correlation-flow-governs-learning-at-criticality-2SecFeedde2026-09-01T17:32:44.213004+00:00Correlation flow governs learning at criticalityhttps://blog.schoergendorfer.com/post/support-selection-beyond-smooth-dag-exactness-completion-geometryscore-margins-a-2SecFeedde2026-09-01T17:32:44.193843+00:00Support Selection Beyond Smooth DAG Exactness: Completion Geometry,Score Margins, and Selective Certificateshttps://blog.schoergendorfer.com/post/crystalgrpo-target-aligned-and-coverage-preserving-reinforcement-learning-for-fl-2SecFeedde2026-09-01T17:32:44.174315+00:00CrystalGRPO: Target-Aligned and Coverage-Preserving Reinforcement Learning for Flow-Based Crystal Structure Predictionhttps://blog.schoergendorfer.com/post/from-uncertainty-to-failure-attribution-self-diagnosing-models-for-failure-attri-2SecFeedde2026-09-01T17:32:44.154620+00:00From Uncertainty to Failure Attribution: Self-Diagnosing Models for Failure Attribution under Distribution Shifthttps://blog.schoergendorfer.com/post/conformalized-large-language-models-under-configuration-shift-2SecFeedde2026-09-01T17:32:44.117676+00:00Conformalized Large Language Models under Configuration Shifthttps://blog.schoergendorfer.com/post/an-identifiability-theory-of-masked-prediction-mode-blindness-and-mask-schedulesSecFeedde2026-09-01T17:17:42.115845+00:00An Identifiability Theory of Masked Prediction: Mode Blindness and Mask Scheduleshttps://blog.schoergendorfer.com/post/learning-what-matters-supervising-global-context-pruning-with-causal-evidence-se-3SecFeedde2026-09-01T17:17:42.072533+00:00Learning What Matters: Supervising Global Context Pruning with Causal Evidence Setshttps://blog.schoergendorfer.com/post/are-single-token-sparse-autoencoder-features-causally-necessary-layer-depth-and-2SecFeedde2026-09-01T17:17:42.049083+00:00Are Single-Token Sparse Autoencoder Features Causally Necessary? Layer-Depth and SAE-Family Effectshttps://blog.schoergendorfer.com/post/spaghetti-architect-a-contamination-resistant-by-construction-labelled-multi-lan-2SecFeedde2026-09-01T17:17:42.028950+00:00Spaghetti Architect: A Contamination-Resistant, By-Construction-Labelled, Multi-Language Code Dataset Generatorhttps://blog.schoergendorfer.com/post/energy-based-physics-informed-form-finding-for-clustered-tensegrity-structures-3SecFeedde2026-09-01T17:17:41.990918+00:00Energy-Based Physics-Informed Form Finding for Clustered Tensegrity Structureshttps://blog.schoergendorfer.com/post/activation-steering-transfer-to-agents-one-gain-ratio-does-not-identify-potency-2SecFeedde2026-09-01T17:17:41.970825+00:00Activation Steering Transfer to Agents: One Gain Ratio Does Not Identify Potency and Efficacyhttps://blog.schoergendorfer.com/post/neureasoner-theory-grounded-mapping-of-reasoning-elicitation-boundariesSecFeedde2026-09-01T17:17:41.937247+00:00NeuReasoner: Theory-grounded Mapping of Reasoning Elicitation Boundarieshttps://blog.schoergendorfer.com/post/wisp-a-working-set-view-of-mixture-of-experts-serving-on-extremely-low-resourceSecFeedde2026-09-01T17:17:41.918457+00:00WiSP: A Working-Set View of Mixture-of-Experts Serving on Extremely Low-Resource Hardwarehttps://blog.schoergendorfer.com/post/learning-generated-controls-under-fractured-geometry-projective-residualizationSecFeedde2026-09-01T17:17:41.897561+00:00Learning Generated Controls under Fractured Geometry: Projective Residualization and Variation-Allocation Frontiershttps://blog.schoergendorfer.com/post/when-design-rules-break-benchmark-composition-determines-whether-label-informatiSecFeedde2026-09-01T17:17:41.861152+00:00When Design Rules Break: Benchmark Composition Determines Whether Label Informativeness Predicts GNN Aggregator Choicehttps://blog.schoergendorfer.com/post/trihead-gan-a-generative-adversarial-network-with-triple-head-discriminator-forSecFeedde2026-09-01T17:17:41.824241+00:00TriHead-GAN: A Generative Adversarial Network with Triple-Head Discriminator for Carbon Emission Time Series Generationhttps://blog.schoergendorfer.com/post/grasp-geometry-aware-residual-alignment-for-scalable-pretraining-data-attributioSecFeedde2026-09-01T17:17:41.803434+00:00GRASP: Geometry-aware Residual Alignment for Scalable Pretraining Data Attributionhttps://blog.schoergendorfer.com/post/mamba-assisted-non-markovian-closure-for-reduced-order-modelingSecFeedde2026-09-01T17:17:41.782016+00:00Mamba-Assisted Non-Markovian Closure for Reduced-Order Modelinghttps://blog.schoergendorfer.com/post/truthful-ai-advisors-a-pre-specified-benchmark-for-large-language-model-honestySecFeedde2026-09-01T17:17:41.759991+00:00Truthful AI Advisors: A Pre-Specified Benchmark for Large Language Model Honesty Under Preference Misalignmenthttps://blog.schoergendorfer.com/post/riemannian-optimization-for-hadamard-products-of-low-rank-matricesSecFeedde2026-09-01T17:02:46.826792+00:00Riemannian Optimization for Hadamard Products of Low-Rank Matriceshttps://blog.schoergendorfer.com/post/generalist-graph-anomaly-detection-via-prototype-based-distillationSecFeedde2026-09-01T17:02:46.805779+00:00Generalist Graph Anomaly Detection via Prototype-Based Distillationhttps://blog.schoergendorfer.com/post/world-model-control-by-trajectory-reachability-metricsSecFeedde2026-09-01T17:02:46.782915+00:00World Model Control by Trajectory Reachability Metricshttps://blog.schoergendorfer.com/post/emulating-the-forced-response-of-climate-models-with-generative-machine-learningSecFeedde2026-09-01T17:02:46.763029+00:00Emulating the Forced Response of Climate Models with Generative Machine Learninghttps://blog.schoergendorfer.com/post/orscale-orthogonalised-optimization-with-layer-wise-trust-ratio-scalingSecFeedde2026-09-01T17:02:46.744074+00:00OrScale: Orthogonalised Optimization with Layer-Wise Trust-Ratio Scalinghttps://blog.schoergendorfer.com/post/admm-q-an-improved-hessian-based-weight-quantizer-for-post-training-quantizationSecFeedde2026-09-01T17:02:46.723302+00:00ADMM-Q: An Improved Hessian-based Weight Quantizer for Post-Training Quantization of Large Language Modelshttps://blog.schoergendorfer.com/post/pairalign-a-framework-for-autoregressive-tokenization-via-self-alignment-with-apSecFeedde2026-09-01T17:02:46.700572+00:00PairAlign: A Framework for Autoregressive Tokenization via Self-Alignment with Applications to Audio Tokenizationhttps://blog.schoergendorfer.com/post/autorec-a-reinforcement-learning-platform-for-equivalent-circuit-model-generatioSecFeedde2026-09-01T17:02:46.680415+00:00AutoREC: A reinforcement learning platform for equivalent circuit model generationhttps://blog.schoergendorfer.com/post/fg2-gdn-enhancing-long-context-gated-delta-networks-with-doubly-fine-grained-conSecFeedde2026-09-01T17:02:46.662268+00:00FG$^2$-GDN: Enhancing Long-Context Gated Delta Networks with Doubly Fine-Grained Controlhttps://blog.schoergendorfer.com/post/realm-reliable-expertise-aware-language-model-fine-tuning-from-noisy-annotationsSecFeedde2026-09-01T17:02:46.638769+00:00REALM: Reliable Expertise-Aware Language Model Fine-Tuning from Noisy Annotationshttps://blog.schoergendorfer.com/post/moonshot-a-framework-for-multi-objective-pruning-of-vision-and-large-language-moSecFeedde2026-09-01T17:02:46.617878+00:00MOONSHOT : A Framework for Multi-Objective Pruning of Vision and Large Language Modelshttps://blog.schoergendorfer.com/post/automated-batch-distillation-process-simulation-for-a-large-hybrid-dataset-for-dSecFeedde2026-09-01T17:02:46.596429+00:00Automated Batch Distillation Process Simulation for a Large Hybrid Dataset for Deep Anomaly Detectionhttps://blog.schoergendorfer.com/post/evolen-evolution-guided-tokenization-for-dna-language-modelSecFeedde2026-09-01T17:02:46.574121+00:00EvoLen: Evolution-Guided Tokenization for DNA Language Modelhttps://blog.schoergendorfer.com/post/libero-para-a-diagnostic-benchmark-and-metrics-for-paraphrase-robustness-in-vlaSecFeedde2026-09-01T17:02:46.548994+00:00LIBERO-Para: A Diagnostic Benchmark and Metrics for Paraphrase Robustness in VLA Modelshttps://blog.schoergendorfer.com/post/identification-of-bivariate-causal-directionality-based-on-anticipated-asymmetriSecFeedde2026-09-01T17:02:46.524473+00:00Identification of Bivariate Causal Directionality Based on Anticipated Asymmetric Geometrieshttps://blog.schoergendorfer.com/post/generalization-and-memorization-in-rectified-flow-2SecFeedde2026-09-01T17:02:46.502693+00:00Generalization and Memorization in Rectified Flowhttps://blog.schoergendorfer.com/post/muse-a-run-centric-platform-for-multimodal-unified-safety-evaluation-of-large-laSecFeedde2026-09-01T17:02:46.457033+00:00MUSE: A Run-Centric Platform for Multimodal Unified Safety Evaluation of Large Language Modelshttps://blog.schoergendorfer.com/post/efficient-real-time-adaptation-of-roms-for-unsteady-flows-using-data-assimilatioSecFeedde2026-09-01T17:02:46.433636+00:00Efficient Real-Time Adaptation of ROMs for Unsteady Flows Using Data Assimilationhttps://blog.schoergendorfer.com/post/constrained-group-relative-policy-optimizationSecFeedde2026-09-01T17:02:46.411489+00:00Constrained Group Relative Policy Optimizationhttps://blog.schoergendorfer.com/post/universal-redundancies-in-time-series-foundation-modelsSecFeedde2026-09-01T16:48:17.347642+00:00Universal Redundancies in Time Series Foundation Modelshttps://blog.schoergendorfer.com/post/federated-personalization-of-early-exit-networksSecFeedde2026-09-01T16:48:17.324085+00:00Federated Personalization of Early-Exit Networkshttps://blog.schoergendorfer.com/post/audit-me-if-you-can-query-efficient-active-fairness-auditing-of-black-box-llmsSecFeedde2026-09-01T16:48:17.304070+00:00Audit Me If You Can: Query-Efficient Active Fairness Auditing of Black-Box LLMshttps://blog.schoergendorfer.com/post/the-double-edged-nature-of-the-rashomon-set-for-trustworthy-machine-learningSecFeedde2026-09-01T16:48:17.268631+00:00The Double-Edged Nature of the Rashomon Set for Trustworthy Machine Learninghttps://blog.schoergendorfer.com/post/deep-reinforcement-learning-for-dynamic-origin-destination-matrix-estimation-in-2SecFeedde2026-09-01T16:48:17.246047+00:00Deep Reinforcement Learning for Dynamic Origin-Destination Matrix Estimation in Microscopic Traffic Simulations Considering Credit Assignmenthttps://blog.schoergendorfer.com/post/structure-preserving-physics-informed-neural-network-for-the-korteweg-de-vries-k-2SecFeedde2026-09-01T16:48:17.225041+00:00Structure-Preserving Physics-Informed Neural Network for the Korteweg--de Vries (KdV) Equationhttps://blog.schoergendorfer.com/post/molga-molecular-graph-adaptation-with-pre-trained-2d-graph-encoder-2SecFeedde2026-09-01T16:48:17.204999+00:00MolGA: Molecular Graph Adaptation with Pre-trained 2D Graph Encoderhttps://blog.schoergendorfer.com/post/multi-marginal-flow-matching-with-adversarially-learnt-interpolantsSecFeedde2026-09-01T16:48:17.183064+00:00Multi-Marginal Flow Matching with Adversarially Learnt Interpolantshttps://blog.schoergendorfer.com/post/data-to-energy-stochastic-dynamicsSecFeedde2026-09-01T16:48:17.136123+00:00Data-to-Energy Stochastic Dynamicshttps://blog.schoergendorfer.com/post/integrating-attention-into-explanation-frameworks-for-language-and-vision-transfSecFeedde2026-09-01T16:48:17.111814+00:00Integrating attention into explanation frameworks for language and vision transformershttps://blog.schoergendorfer.com/post/agnostics-learning-to-code-in-any-programming-language-via-reinforcement-with-aSecFeedde2026-09-01T16:48:17.088120+00:00Agnostics: Learning to Code in Any Programming Language via Reinforcement with a Universal Learning Environmenthttps://blog.schoergendorfer.com/post/a-conditional-gan-for-tabular-data-generation-with-probabilistic-sampling-of-latSecFeedde2026-09-01T16:48:17.061671+00:00A Conditional GAN for Tabular Data Generation with Probabilistic Sampling of Latent Subspaceshttps://blog.schoergendorfer.com/post/a-cycle-consistency-constrained-framework-for-dynamic-solution-space-reduction-iSecFeedde2026-09-01T16:48:17.036527+00:00A Cycle-Consistency Constrained Framework for Dynamic Solution Space Reduction in Noninjective Regressionhttps://blog.schoergendorfer.com/post/discrete-compositional-generation-via-general-soft-operators-and-robust-reinforcSecFeedde2026-09-01T16:48:17.015694+00:00Discrete Compositional Generation via General Soft Operators and Robust Reinforcement Learninghttps://blog.schoergendorfer.com/post/semantics-at-an-angle-when-cosine-similarity-works-until-it-doesnt-2SecFeedde2026-09-01T16:32:40.523861+00:00Semantics at an Angle: When Cosine Similarity Works Until It Doesn'thttps://blog.schoergendorfer.com/post/temporal-analysis-of-netflow-datasets-for-network-intrusion-detection-systemsSecFeedde2026-09-01T16:32:40.499545+00:00Temporal Analysis of NetFlow Datasets for Network Intrusion Detection Systemshttps://blog.schoergendorfer.com/post/h-fedsn-personalized-sparse-networks-for-efficient-and-accurate-hierarchical-fedSecFeedde2026-09-01T16:32:40.441771+00:00H-FedSN: Personalized Sparse Networks for Efficient and Accurate Hierarchical Federated Learning for IoT Applicationshttps://blog.schoergendorfer.com/post/mega-message-passing-neural-networks-for-multigraphs-with-edge-attributesSecFeedde2026-09-01T16:32:40.418691+00:00MEGA: Message Passing Neural Networks for Multigraphs with EdGe Attributeshttps://blog.schoergendorfer.com/post/adaptive-teachers-for-amortized-samplersSecFeedde2026-09-01T16:32:40.395891+00:00Adaptive teachers for amortized samplershttps://blog.schoergendorfer.com/post/delta-ai-local-objectives-for-amortized-inference-in-sparse-graphical-modelsSecFeedde2026-09-01T16:32:40.311791+00:00Delta-AI: Local objectives for amortized inference in sparse graphical modelshttps://blog.schoergendorfer.com/post/implementing-neural-network-mixed-effects-models-in-template-model-builder-tmbSecFeedde2026-09-01T16:32:40.274411+00:00Implementing neural network mixed-effects models in Template Model Builder (TMB)https://blog.schoergendorfer.com/post/one-adapter-many-tasks-task-conditioned-feature-transformations-for-continual-leSecFeedde2026-09-01T16:32:40.252914+00:00One Adapter, Many Tasks: Task-Conditioned Feature Transformations for Continual Learninghttps://blog.schoergendorfer.com/post/train-classical-deploy-quantum-requires-rethinking-generalizationSecFeedde2026-09-01T16:32:40.229096+00:00"Train classical, deploy quantum" requires rethinking generalizationhttps://blog.schoergendorfer.com/post/minimax-bounds-for-watermarked-and-masked-recursive-discrete-distribution-estimaSecFeedde2026-09-01T16:32:40.205890+00:00Minimax bounds for watermarked and masked recursive discrete distribution estimationhttps://blog.schoergendorfer.com/post/segmentation-of-bovid-dentition-under-imperfect-annotations-a-comparative-studySecFeedde2026-09-01T16:32:40.182011+00:00Segmentation of Bovid Dentition Under Imperfect Annotations: A Comparative Study of Convolutional and Attention Modelshttps://blog.schoergendorfer.com/post/driving-on-memorySecFeedde2026-09-01T16:32:40.157414+00:00Driving on Memoryhttps://blog.schoergendorfer.com/post/learning-the-geometry-of-admissible-hypotheses-through-inductive-bias-in-traininSecFeedde2026-09-01T16:32:40.129925+00:00Learning the Geometry of Admissible Hypotheses through Inductive Bias in Training Distributionshttps://blog.schoergendorfer.com/post/uncertainty-aware-end-to-end-ai-weather-forecasting-disentangling-observation-anSecFeedde2026-09-01T16:17:56.429268+00:00Uncertainty-Aware End-to-End AI Weather Forecasting: Disentangling Observation and Model Contributionshttps://blog.schoergendorfer.com/post/what-it-costs-to-compose-rebuild-and-correct-precomputed-memorySecFeedde2026-09-01T16:17:56.410513+00:00What It Costs to Compose, Rebuild, and Correct Precomputed Memoryhttps://blog.schoergendorfer.com/post/quantum-grassmann-plucker-token-mixing-for-deep-learning-based-post-disaster-damSecFeedde2026-09-01T16:17:56.391251+00:00Quantum-Grassmann-Plucker Token Mixing for Deep Learning-Based Post-Disaster Damage Assessmenthttps://blog.schoergendorfer.com/post/informative-label-missingness-in-multiclass-classification-information-geometrySecFeedde2026-09-01T16:17:56.371188+00:00Informative Label Missingness in Multiclass Classification Information Geometry and Excess Riskhttps://blog.schoergendorfer.com/post/minerals-in-the-wild-a-hyperspectral-xrf-dataset-for-elemental-composition-estimSecFeedde2026-09-01T16:17:56.351695+00:00Minerals in the Wild: A Hyperspectral-XRF Dataset for Elemental Composition Estimationhttps://blog.schoergendorfer.com/post/two-centuries-of-sexism-in-british-parliament-a-computational-analysis-of-womensSecFeedde2026-09-01T16:17:56.315585+00:00Two Centuries of Sexism in British Parliament: A Computational Analysis of Women's Representation in the Hansard Corpushttps://blog.schoergendorfer.com/post/viser-visual-evidence-and-reliance-for-object-hallucination-detection-in-lvlmsSecFeedde2026-09-01T16:17:56.294920+00:00VisER: Visual Evidence and Reliance for Object Hallucination Detection in LVLMshttps://blog.schoergendorfer.com/post/end-to-end-neural-shrinkage-of-indefinite-pairwise-correlation-matrices-for-smalSecFeedde2026-09-01T16:17:56.272013+00:00End-to-End Neural Shrinkage of Indefinite Pairwise Correlation Matrices for Small-Cap-Inclusive Portfolioshttps://blog.schoergendorfer.com/post/event-driven-language-models-with-sparse-neural-activity-for-neuromorphic-hardwaSecFeedde2026-09-01T16:17:56.248627+00:00Event-Driven Language Models with Sparse Neural Activity for Neuromorphic Hardwarehttps://blog.schoergendorfer.com/post/ceiling-clipped-acceptance-histograms-indicate-stranded-speed-up-in-block-diffusSecFeedde2026-09-01T16:17:56.208131+00:00Ceiling-Clipped Acceptance Histograms Indicate Stranded Speed-up in Block-Diffusion Speculative Decodinghttps://blog.schoergendorfer.com/post/benchmarking-external-generalization-of-spd-matrix-learning-for-resting-state-fmSecFeedde2026-09-01T16:17:56.180787+00:00Benchmarking External Generalization of SPD Matrix Learning for Resting-State fMRI Connectome Predictionhttps://blog.schoergendorfer.com/post/kathleen-remembers-length-invariant-one-shot-recall-without-attentionSecFeedde2026-09-01T16:17:56.157289+00:00Kathleen Remembers: Length-Invariant One-Shot Recall Without Attentionhttps://blog.schoergendorfer.com/post/compact-and-infinite-order-error-analysis-for-null-space-svd-estimationSecFeedde2026-09-01T16:17:56.133548+00:00Compact and Infinite-Order Error Analysis for Null-Space SVD Estimationhttps://blog.schoergendorfer.com/post/strengthening-recursive-constructions-for-zero-error-shannon-capacitySecFeedde2026-09-01T16:17:56.109852+00:00Strengthening Recursive Constructions for Zero-Error Shannon Capacityhttps://blog.schoergendorfer.com/post/a-borel-concept-class-of-vc-dimension-one-with-a-non-pac-consistent-learner-in-zSecFeedde2026-09-01T16:17:56.068443+00:00A Borel Concept Class of VC Dimension One with a Non-PAC Consistent Learner in ZFChttps://blog.schoergendorfer.com/post/fairness-in-multi-class-multi-group-classification-problems-via-contextial-coherSecFeedde2026-09-01T16:17:56.045158+00:00Fairness in multi-class multi-group classification problems via contextial coherent risk measureshttps://blog.schoergendorfer.com/post/the-pur-1-cyber-physical-digital-twinSecFeedde2026-09-01T16:03:49.631531+00:00The PUR-1 Cyber-Physical Digital Twinhttps://blog.schoergendorfer.com/post/balancing-privacy-utility-and-safety-in-llm-alignment-through-preference-optimizSecFeedde2026-09-01T16:03:49.614112+00:00Balancing Privacy, Utility, and Safety in LLM Alignment through Preference Optimizationhttps://blog.schoergendorfer.com/post/a-hybrid-state-space-approach-for-census-tract-population-estimationSecFeedde2026-09-01T16:03:49.575144+00:00A Hybrid State-Space Approach for Census-Tract Population Estimationhttps://blog.schoergendorfer.com/post/when-does-a-classifier-help-an-llm-classifier-guided-prompting-and-hybrid-classiSecFeedde2026-09-01T16:03:49.557352+00:00When Does a Classifier Help an LLM? Classifier-Guided Prompting and Hybrid Classifier-LLM Models for Credit-Default Predictionhttps://blog.schoergendorfer.com/post/learning-representations-through-token-prediction-geometry-approximation-and-dowSecFeedde2026-09-01T16:03:49.535603+00:00Learning Representations through Token Prediction: Geometry, Approximation, and Downstream Guaranteeshttps://blog.schoergendorfer.com/post/a-deep-latent-variable-framework-for-jointly-modeling-missingness-measurement-erSecFeedde2026-09-01T16:03:49.496764+00:00A Deep Latent Variable Framework for Jointly Modeling Missingness, Measurement Error, and Heterogeneityhttps://blog.schoergendorfer.com/post/data-driven-design-optimization-of-streaming-potential-mediated-electrokinetic-tSecFeedde2026-09-01T16:03:49.472873+00:00Data-Driven Design Optimization of Streaming-Potential-Mediated Electrokinetic Transport of Viscoelastic Fluids in Microchannelshttps://blog.schoergendorfer.com/post/when-safety-speaks-a-language-a-mechanistic-analysis-of-safety-language-identitySecFeedde2026-09-01T16:03:49.449549+00:00When Safety Speaks a Language: A Mechanistic Analysis of Safety-Language Identity Entanglement in LLMshttps://blog.schoergendorfer.com/post/compression-aware-abstention-teaching-llms-to-refuse-when-kv-compression-masks-rSecFeedde2026-09-01T16:03:49.425701+00:00Compression-Aware Abstention: Teaching LLMs to Refuse When KV-Compression Masks Remove Answer Evidencehttps://blog.schoergendorfer.com/post/towards-continual-test-time-adaptation-of-vision-language-models-in-open-vocabulSecFeedde2026-09-01T16:03:49.405616+00:00Towards Continual Test-Time Adaptation of Vision-Language Models in Open-Vocabulary Semantic Segmentationhttps://blog.schoergendorfer.com/post/the-price-of-intelligence-a-quality-adjusted-price-index-for-ai-servicesSecFeedde2026-09-01T16:03:49.366854+00:00The Price of Intelligence: A Quality-Adjusted Price Index for AI Serviceshttps://blog.schoergendorfer.com/post/influence-directed-distillation-solving-the-diversity-bottleneck-in-sampled-tokeSecFeedde2026-09-01T16:03:49.345008+00:00Influence-Directed Distillation: Solving the Diversity Bottleneck in Sampled-Token On-Policy Distillationhttps://blog.schoergendorfer.com/post/a-unified-perspective-on-conformal-prediction-and-wasserstein-distributionally-rSecFeedde2026-09-01T16:03:49.326186+00:00A Unified Perspective on Conformal Prediction and Wasserstein Distributionally Robust Optimization for Uncertainty Quantificationhttps://blog.schoergendorfer.com/post/neural-ode-enhanced-linear-mixed-effect-models-for-estimating-complex-associatioSecFeedde2026-09-01T16:03:49.310047+00:00Neural ODE enhanced linear mixed effect models for estimating complex association patterns of time-varying covariates with the marker trajectoryhttps://blog.schoergendorfer.com/post/transformer-based-flow-shop-scheduling-using-milp-generated-training-dataSecFeedde2026-09-01T16:03:49.291478+00:00Transformer-Based Flow Shop Scheduling Using MILP-Generated Training Datahttps://blog.schoergendorfer.com/post/buttermamba-butterworth-enhanced-spatial-temporal-mamba-for-efficient-traffic-flSecFeedde2026-09-01T16:03:49.269657+00:00ButterMamba: Butterworth-Enhanced Spatial-Temporal Mamba for Efficient Traffic Flow Predictionhttps://blog.schoergendorfer.com/post/cross-lingual-functional-vectors-for-emotion-detection-in-large-language-modelsSecFeedde2026-09-01T16:03:49.246017+00:00Cross-lingual Functional Vectors for Emotion Detection in Large Language Modelshttps://blog.schoergendorfer.com/post/mathcaln_0-foundation-towards-the-age-of-tactile-intelligenceSecFeedde2026-09-01T15:47:52.422041+00:00$\mathcal{N}_0$-Foundation: Towards the Age of Tactile Intelligencehttps://blog.schoergendorfer.com/post/tacs-trajectory-aware-candidate-selection-for-llm-jailbreak-suffix-optimizationSecFeedde2026-09-01T15:47:52.401360+00:00TACS: Trajectory-Aware Candidate Selection for LLM Jailbreak Suffix Optimizationhttps://blog.schoergendorfer.com/post/logo-token-level-dynamic-local-global-attentionSecFeedde2026-09-01T15:47:52.379907+00:00LoGo: Token-Level Dynamic Local-Global Attentionhttps://blog.schoergendorfer.com/post/armor-manifold-oriented-training-for-adversarially-robust-aerial-object-detectioSecFeedde2026-09-01T15:47:52.362592+00:00ARMOR: Manifold-Oriented Training for Adversarially Robust Aerial Object Detection under Data Scarcityhttps://blog.schoergendorfer.com/post/deciding-when-to-decide-testing-operational-suboptimality-under-distributional-sSecFeedde2026-09-01T15:47:52.343679+00:00Deciding When to Decide: Testing Operational Suboptimality Under Distributional Shifthttps://blog.schoergendorfer.com/post/item-mean-surrogates-why-richer-persona-data-fail-to-improve-llms-as-human-surroSecFeedde2026-09-01T15:47:52.321552+00:00Item-Mean Surrogates: Why Richer Persona Data Fail to Improve LLMs as Human Surrogateshttps://blog.schoergendorfer.com/post/content-exploration-beyond-the-feed-creator-supply-and-the-shared-corpusSecFeedde2026-09-01T15:47:52.300979+00:00Content Exploration Beyond the Feed: Creator Supply and the Shared Corpushttps://blog.schoergendorfer.com/post/film-gpnet-geometry-aware-pseudo-supervised-phase-restoration-with-zero-shot-genSecFeedde2026-09-01T15:47:52.245159+00:00FiLM-GPNet: Geometry-Aware Pseudo-Supervised Phase Restoration with Zero-Shot Generalization for Large Temporal InSAR Stackshttps://blog.schoergendorfer.com/post/spectral-analysis-for-sparse-matrix-computation-insights-and-potentialSecFeedde2026-09-01T15:47:52.225150+00:00Spectral Analysis for Sparse Matrix Computation: Insights and Potentialhttps://blog.schoergendorfer.com/post/signed-random-fourier-features-for-fast-density-estimation-with-indefinite-kerneSecFeedde2026-09-01T15:47:52.205188+00:00Signed random Fourier features for fast density estimation with indefinite kernelshttps://blog.schoergendorfer.com/post/sense-once-serve-many-common-trace-factorized-constrained-ppo-for-online-sensingSecFeedde2026-09-01T15:47:52.184478+00:00Sense Once, Serve Many: Common-Trace Factorized Constrained PPO for Online Sensing-Session Consolidation in Multi-Tenant ISAC Networkshttps://blog.schoergendorfer.com/post/adavla-adaptive-step-flow-matching-for-training-free-acceleration-of-vision-langSecFeedde2026-09-01T15:47:52.167545+00:00AdaVLA: Adaptive Step Flow Matching for Training-free Acceleration of Vision-Language-Action Modelshttps://blog.schoergendorfer.com/post/a-madiff-attention-guided-multi-agent-drl-with-diffusion-policies-for-memory-awaSecFeedde2026-09-01T15:47:52.148093+00:00A-MADiff: Attention-Guided Multi-Agent DRL with Diffusion Policies for Memory-Aware Task Orchestration in Mobile AIGC Networkshttps://blog.schoergendorfer.com/post/uniform-statistical-convergence-of-empirical-sinkhorn-potentials-with-exponentiaSecFeedde2026-09-01T15:47:52.127585+00:00Uniform Statistical Convergence of Empirical Sinkhorn Potentials with Exponential and Polynomial Dependence on the Regularization Parameterhttps://blog.schoergendorfer.com/post/optimally-selecting-representative-agents-from-a-metric-spaceSecFeedde2026-09-01T15:47:52.104477+00:00Optimally Selecting Representative Agents from a Metric Spacehttps://blog.schoergendorfer.com/post/spectral-embedded-operator-learning-for-three-phase-interfacial-flow-a-ternary-cSecFeedde2026-09-01T15:47:52.080305+00:00Spectral-Embedded Operator Learning for Three-Phase Interfacial Flow: A Ternary Cahn-Hilliard-Navier-Stokes Benchmarkhttps://blog.schoergendorfer.com/post/jigsaw-crl-recovering-global-latent-causal-order-from-fragmented-multi-client-inSecFeedde2026-09-01T15:47:52.037805+00:00Jigsaw-CRL: Recovering Global Latent Causal Order from Fragmented Multi-Client Interventionshttps://blog.schoergendorfer.com/post/brain-language-action-bla-models-language-conditioned-eeg-for-robotics-controlSecFeedde2026-09-01T15:33:21.810216+00:00Brain-Language-Action (BLA) Models: Language-Conditioned EEG for Robotics Controlhttps://blog.schoergendorfer.com/post/merit-mitigating-exposure-bias-in-generative-xmc-for-user-interest-propensity-moSecFeedde2026-09-01T15:33:21.785234+00:00MERIT: Mitigating Exposure Bias in Generative XMC for User-Interest Propensity Modelinghttps://blog.schoergendorfer.com/post/mmir-frequency-space-inverse-rendering-for-3d-millimeter-wave-radar-adc-synthesiSecFeedde2026-09-01T15:33:21.735866+00:00mmIR: Frequency-Space Inverse Rendering for 3D Millimeter-Wave Radar ADC Synthesishttps://blog.schoergendorfer.com/post/mwir-4-plastic-the-identification-of-complex-end-of-life-industrial-plastic-usinSecFeedde2026-09-01T15:33:21.712479+00:00MWIR-4-Plastic: The Identification of Complex End-of-Life Industrial Plastic using Mid-wave Infrared Hyperspectral Imaging and Machine Learninghttps://blog.schoergendorfer.com/post/generative-translation-priors-bayesian-imaging-with-cross-modality-image-translaSecFeedde2026-09-01T15:33:21.686624+00:00Generative Translation Priors: Bayesian Imaging with Cross-Modality Image Translationhttps://blog.schoergendorfer.com/post/quantitative-target-convergence-and-uniform-in-time-propagation-of-chaos-for-lanSecFeedde2026-09-01T15:33:21.666298+00:00Quantitative Target Convergence and Uniform-in-Time Propagation of Chaos for Langevin-Regularized SVGDhttps://blog.schoergendorfer.com/post/separable-nonnegative-matrix-factorization-using-powered-ratio-of-norms-regulariSecFeedde2026-09-01T15:33:21.623337+00:00Separable Nonnegative Matrix Factorization Using Powered Ratio-of-Norms Regularizationhttps://blog.schoergendorfer.com/post/adversarial-calibration-attack-on-autonomous-vehiclesSecFeedde2026-09-01T15:33:21.602364+00:00Adversarial Calibration Attack on Autonomous Vehicleshttps://blog.schoergendorfer.com/post/generation-of-high-level-concepts-in-3d-scene-graphs-via-autoregressive-diffusioSecFeedde2026-09-01T15:33:21.581219+00:00Generation of High-Level Concepts in 3D Scene Graphs via Autoregressive Diffusionhttps://blog.schoergendorfer.com/post/data-diversity-not-frequency-invariance-a-controlled-and-self-audited-study-of-cSecFeedde2026-09-01T15:33:21.561066+00:00Data Diversity, Not Frequency Invariance: A Controlled and Self-Audited Study of Compression-Robust Deepfake Detectionhttps://blog.schoergendorfer.com/post/framescope-temporal-data-valuation-for-stream-active-learning-in-autonomous-vehiSecFeedde2026-09-01T15:33:21.507367+00:00FrameScope: Temporal Data Valuation for Stream Active Learning in Autonomous Vehicle Systemshttps://blog.schoergendorfer.com/post/propensity-straight-through-gradients-for-discrete-stochastic-systemsSecFeedde2026-09-01T15:33:21.486954+00:00Propensity Straight-Through Gradients for Discrete Stochastic Systemshttps://blog.schoergendorfer.com/post/privacy-preserving-detection-of-rare-disease-associated-cell-subsets-via-secureSecFeedde2026-09-01T15:33:21.467633+00:00Privacy-Preserving Detection of Rare Disease-Associated Cell Subsets via Secure Multi-Party Computationhttps://blog.schoergendorfer.com/post/sharp-approximation-rates-for-neural-networks-with-affine-latent-parameterizatioSecFeedde2026-09-01T15:33:21.431313+00:00Sharp Approximation Rates for Neural Networks with Affine Latent Parameterizationshttps://blog.schoergendorfer.com/post/on-the-complexity-of-the-compatibility-problem-for-succinctly-encoded-conditionaSecFeedde2026-09-01T15:33:21.411361+00:00On the Complexity of the Compatibility Problem for Succinctly Encoded Conditional Distributionshttps://blog.schoergendorfer.com/post/stress-testing-efficient-responsible-ai-evaluation-when-compute-savings-change-bSecFeedde2026-09-01T15:17:53.356635+00:00Stress-Testing Efficient Responsible-AI Evaluation: When Compute Savings Change Benchmark Conclusionshttps://blog.schoergendorfer.com/post/sycophantic-agreement-transfers-with-neutral-data-via-contrastive-preference-optSecFeedde2026-09-01T15:17:53.336016+00:00Sycophantic Agreement Transfers with Neutral Data via Contrastive Preference Optimizationhttps://blog.schoergendorfer.com/post/universal-transformers-for-circuit-computations-perfect-length-generalization-inSecFeedde2026-09-01T15:17:53.296518+00:00Universal Transformers for Circuit Computations: Perfect Length Generalization in Tiny Transformershttps://blog.schoergendorfer.com/post/does-on-policy-distillation-really-distill-from-noisy-teacher-to-self-improvemenSecFeedde2026-09-01T15:17:53.275246+00:00Does On-Policy Distillation Really Distill? From Noisy Teacher to Self-Improvementhttps://blog.schoergendorfer.com/post/rotational-equivariance-in-machine-learning-a-comprehensive-tutorialSecFeedde2026-09-01T15:17:53.251519+00:00Rotational Equivariance in Machine Learning: A Comprehensive Tutorialhttps://blog.schoergendorfer.com/post/tspfn-a-temporal-tabular-foundation-model-for-physiological-time-series-classifiSecFeedde2026-09-01T15:17:53.218124+00:00TSPFN: A Temporal Tabular Foundation Model for Physiological Time Series Classificationhttps://blog.schoergendorfer.com/post/language-informed-flow-matching-for-trend-guided-structure-based-3d-molecular-geSecFeedde2026-09-01T15:17:53.198775+00:00Language-Informed Flow Matching for Trend-Guided Structure-Based 3D Molecular Generationhttps://blog.schoergendorfer.com/post/sparse-competition-during-training-for-the-emergence-of-specialized-modulesSecFeedde2026-09-01T15:17:53.159099+00:00Sparse Competition during Training For the Emergence of Specialized Moduleshttps://blog.schoergendorfer.com/post/a-human-in-the-loop-autonomous-agent-for-industry-time-series-forecastingSecFeedde2026-09-01T15:17:53.137446+00:00A Human-in-the-Loop Autonomous Agent for Industry Time Series Forecastinghttps://blog.schoergendorfer.com/post/singular-curvature-in-relu-trainingdifferentiation-and-the-gradient-flow-limit-nSecFeedde2026-09-01T15:17:53.116659+00:00Singular Curvature in ReLU Training:Differentiation and the Gradient-Flow Limit Need Not Commutehttps://blog.schoergendorfer.com/post/one-policy-is-enough-single-agent-reinforcement-learning-outperforms-tree-searchSecFeedde2026-09-01T15:17:53.098247+00:00One Policy Is Enough: Single-Agent Reinforcement Learning Outperforms Tree Search for Chemistry Tool Learninghttps://blog.schoergendorfer.com/post/reproducible-macroscopic-dynamics-in-a-closed-loop-human-ai-learning-systemSecFeedde2026-09-01T15:17:53.071389+00:00Reproducible macroscopic dynamics in a closed-loop human-AI learning systemhttps://blog.schoergendorfer.com/post/nonparametric-contextual-pricing-and-inventory-learning-under-censored-demandSecFeedde2026-09-01T15:17:53.045571+00:00Nonparametric Contextual Pricing and Inventory Learning under Censored Demandhttps://blog.schoergendorfer.com/post/s3c-llm-skill-code-guided-agentic-language-models-for-spectrum-to-structure-elucSecFeedde2026-09-01T15:17:53.008781+00:00S3C-LLM: Skill-Code Guided Agentic Language Models for Spectrum-to-Structure Elucidationhttps://blog.schoergendorfer.com/post/deploying-deepseek-175b-locally-on-a-single-consumer-grade-rtx-4060-laptop-withSecFeedde2026-09-01T15:17:52.969456+00:00Deploying DeepSeek 175B Locally on a Single Consumer-Grade RTX 4060 Laptop with 32GB RAM for 200k-Scale Protein-Ligand Virtual Screeninghttps://blog.schoergendorfer.com/post/geometric-attractor-monitoring-a-robust-and-frugal-framework-for-multi-modal-indSecFeedde2026-09-01T15:03:44.852616+00:00Geometric Attractor Monitoring: A Robust and Frugal Framework for Multi-modal Industrial Robotic Cycleshttps://blog.schoergendorfer.com/post/reciprocity-separates-gradient-flow-from-rotation-in-conservative-physical-learnSecFeedde2026-09-01T15:03:44.829301+00:00Reciprocity Separates Gradient Flow from Rotation in Conservative Physical Learninghttps://blog.schoergendorfer.com/post/t3s-improving-multi-task-reinforcement-learning-with-task-specific-feature-selecSecFeedde2026-09-01T15:03:44.785991+00:00T3S: Improving Multi-Task Reinforcement Learning with Task-Specific Feature Selector and Schedulerhttps://blog.schoergendorfer.com/post/practice-from-experience-to-expertise-in-self-evolving-embodied-agentsSecFeedde2026-09-01T15:03:44.761373+00:00PRACTICE: From Experience to Expertise in Self-Evolving Embodied Agentshttps://blog.schoergendorfer.com/post/do-vlms-share-safety-neurons-across-modalitiesSecFeedde2026-09-01T15:03:44.743533+00:00Do VLMs Share Safety Neurons Across Modalities?https://blog.schoergendorfer.com/post/tddm-melatt-a-decoupled-memory-and-diffusion-framework-for-generalizable-encryptSecFeedde2026-09-01T15:03:44.715288+00:00TDDM-Melatt: A Decoupled Memory and Diffusion Framework for Generalizable Encrypted Traffic Classificationhttps://blog.schoergendorfer.com/post/e-commerce-bench-evaluating-llm-agents-on-long-horizon-autonomous-business-operaSecFeedde2026-09-01T15:03:44.681782+00:00E-Commerce Bench: Evaluating LLM Agents on Long-Horizon Autonomous Business Operationhttps://blog.schoergendorfer.com/post/tracing-distinguishability-through-transformer-processing-with-stochastic-layernSecFeedde2026-09-01T15:03:44.662015+00:00Tracing distinguishability through transformer processing with stochastic LayerNormhttps://blog.schoergendorfer.com/post/liquid-gated-attentionSecFeedde2026-09-01T15:03:44.625407+00:00Liquid Gated Attentionhttps://blog.schoergendorfer.com/post/season-aware-hybrid-convolutional-transformer-for-antarctic-sea-ice-concentratioSecFeedde2026-09-01T15:03:44.604049+00:00Season-Aware Hybrid Convolutional-Transformer for Antarctic Sea Ice Concentration Forecastinghttps://blog.schoergendorfer.com/post/three-steps-at-a-time-learning-representations-from-action-sequences-in-contrastSecFeedde2026-09-01T15:03:44.579522+00:00Three Steps at a Time: Learning Representations from Action Sequences in Contrastive RLhttps://blog.schoergendorfer.com/post/state-of-health-estimation-using-convolutional-and-bidirectional-lstm-neural-netSecFeedde2026-09-01T15:03:44.523140+00:00State of Health Estimation using Convolutional and Bidirectional LSTM Neural Networks tuned by Bayesian Optimizationhttps://blog.schoergendorfer.com/post/the-safety-relay-in-roleplay-jailbreaks-a-component-resolved-causal-analysis-ofSecFeedde2026-09-01T15:03:44.500674+00:00The Safety Relay in Roleplay Jailbreaks: A Component-Resolved Causal Analysis of Harm Recognition and Refusalhttps://blog.schoergendorfer.com/post/pac-progress-augmented-advantage-curriculum-for-multi-task-reinforcement-learninSecFeedde2026-09-01T15:03:44.475517+00:00PAC: Progress-Augmented Advantage Curriculum for Multi-Task Reinforcement Learning of LLMshttps://blog.schoergendorfer.com/post/when-the-martingale-never-stops-firing-anytime-valid-gating-on-real-forecast-strSecFeedde2026-09-01T14:48:52.369274+00:00When the Martingale Never Stops Firing: Anytime-Valid Gating on Real Forecast Streamshttps://blog.schoergendorfer.com/post/toxlens-a-reproducible-graph-learning-framework-for-leakage-aware-uncertainty-caSecFeedde2026-09-01T14:48:52.349165+00:00ToxLens: A Reproducible Graph-Learning Framework for Leakage-Aware, Uncertainty-Calibrated Molecular Toxicity Predictionhttps://blog.schoergendorfer.com/post/learning-where-outcomes-changecredit-addressable-reasoning-for-multimodal-geometSecFeedde2026-09-01T14:48:52.328219+00:00Learning Where Outcomes Change:Credit-Addressable Reasoning for Multimodal Geometryhttps://blog.schoergendorfer.com/post/self-supervised-pretext-tasks-for-infant-cry-analysis-a-controlled-comparison-anSecFeedde2026-09-01T14:48:52.306800+00:00Self-Supervised Pretext Tasks for Infant Cry Analysis: A Controlled Comparison and a Cautionary Result on Donateacryhttps://blog.schoergendorfer.com/post/prime-mitigating-subgroup-optimization-competition-in-shared-ctr-top-networks-wiSecFeedde2026-09-01T14:48:52.285543+00:00PRIME: Mitigating Subgroup Optimization Competition in Shared CTR Top Networks with Plug-in Residual Input-Conditioned Mixture of Experthttps://blog.schoergendorfer.com/post/confounding-masquerading-as-improvement-a-systematic-evaluation-of-offline-reinfSecFeedde2026-09-01T14:48:52.265207+00:00Confounding Masquerading as Improvement: A Systematic Evaluation of Offline Reinforcement Learning for Stroke Antithrombotic Treatment in a 129,000-Patient Registryhttps://blog.schoergendorfer.com/post/locally-guided-actor-critic-training-a-goal-conditioned-actor-with-a-subgoal-awaSecFeedde2026-09-01T14:48:52.230248+00:00Locally-Guided Actor-Critic: Training a Goal-conditioned Actor with a Subgoal-aware Critichttps://blog.schoergendorfer.com/post/foundation-models-meet-agriculture-challenges-beyond-pretrainingSecFeedde2026-09-01T14:48:52.209114+00:00Foundation Models Meet Agriculture: Challenges Beyond Pretraininghttps://blog.schoergendorfer.com/post/uncertainty-of-vision-medical-foundation-modelsSecFeedde2026-09-01T14:48:52.189587+00:00Uncertainty of Vision Medical Foundation Modelshttps://blog.schoergendorfer.com/post/rslm-training-free-vector-quantization-for-approximate-nearest-neighbor-searchSecFeedde2026-09-01T14:48:52.170478+00:00RSLM: Training-Free Vector Quantization for Approximate Nearest Neighbor Searchhttps://blog.schoergendorfer.com/post/convergence-rates-for-the-rmsprop-optimizer-with-full-control-of-the-hyperparameSecFeedde2026-09-01T14:48:52.150510+00:00Convergence rates for the RMSprop optimizer with full control of the hyperparametershttps://blog.schoergendorfer.com/post/beat-synchronous-tokenization-for-ecg-transformersSecFeedde2026-09-01T14:48:52.129087+00:00Beat-Synchronous Tokenization for ECG Transformershttps://blog.schoergendorfer.com/post/beyond-churn-predicting-financial-fragmentation-in-retail-banking-with-temporalSecFeedde2026-09-01T14:48:52.086894+00:00Beyond Churn: Predicting Financial Fragmentation in Retail Banking with Temporal Machine Learninghttps://blog.schoergendorfer.com/post/coarse-composition-suffices-tabular-in-context-learning-for-multi-activity-antimSecFeedde2026-09-01T14:48:52.065988+00:00Coarse composition suffices: tabular in-context learning for multi-activity antimicrobial peptide profilinghttps://blog.schoergendorfer.com/post/generative-multi-domain-transfer-learning-for-fault-detection-in-data-scarce-winSecFeedde2026-09-01T14:48:52.047074+00:00Generative multi-domain transfer learning for fault detection in data-scarce wind turbineshttps://blog.schoergendorfer.com/post/learning-pde-time-stepping-with-neural-cellular-automataSecFeedde2026-09-01T14:48:52.028895+00:00Learning PDE Time-Stepping with Neural Cellular Automatahttps://blog.schoergendorfer.com/post/context-staircase-signature-aligned-dynamics-of-token-embeddings-under-small-iniSecFeedde2026-09-01T14:48:52.006802+00:00Context Staircase: Signature-Aligned Dynamics of Token Embeddings under Small Initializationhttps://blog.schoergendorfer.com/post/multivariate-scientific-data-compression-with-learned-cross-variable-latent-decoSecFeedde2026-09-01T14:48:51.984908+00:00Multivariate Scientific Data Compression with Learned Cross-Variable Latent Decorrelation and Autoregressive Entropy Modelinghttps://blog.schoergendorfer.com/post/exact-recovery-thresholds-for-weighted-data-selection-in-vector-valued-linear-reSecFeedde2026-09-01T14:33:13.568268+00:00Exact Recovery Thresholds for Weighted Data Selection in Vector-Valued Linear Regressionhttps://blog.schoergendorfer.com/post/strong-drafts-need-compact-memories-long-context-speculative-decoding-with-comprSecFeedde2026-09-01T14:33:13.545055+00:00Strong Drafts Need Compact Memories: Long-Context Speculative Decoding with Compressed KV Cachehttps://blog.schoergendorfer.com/post/diffusion-based-refinement-for-kilometer-scale-probabilistic-precipitation-nowcaSecFeedde2026-09-01T14:33:13.527074+00:00Diffusion-Based Refinement for Kilometer-Scale Probabilistic Precipitation Nowcastinghttps://blog.schoergendorfer.com/post/certified-safety-radii-in-forecast-error-space-for-wasserstein-distributionallySecFeedde2026-09-01T14:33:13.506718+00:00Certified Safety Radii in Forecast-Error Space for Wasserstein Distributionally Robust Small Signal Stability-Constrained AC Optimal Power Flow via Lifted Spectrahedral Containmenthttps://blog.schoergendorfer.com/post/benchmarking-peptide-protein-affinity-prediction-across-peptide-and-target-shiftSecFeedde2026-09-01T14:33:13.483009+00:00Benchmarking Peptide-Protein Affinity Prediction Across Peptide and Target Shiftshttps://blog.schoergendorfer.com/post/reinforcement-learning-for-symbolic-equation-solvingSecFeedde2026-09-01T14:33:13.452436+00:00Reinforcement Learning for Symbolic Equation Solvinghttps://blog.schoergendorfer.com/post/supraglacial-lake-fate-is-knowable-long-before-the-season-endsSecFeedde2026-09-01T14:33:13.415710+00:00Supraglacial Lake Fate Is Knowable Long Before the Season Endshttps://blog.schoergendorfer.com/post/smote-var-an-uncertainty-aware-oversampling-method-for-predicting-depression-remSecFeedde2026-09-01T14:33:13.393277+00:00SMOTE-VAR: An Uncertainty-Aware Oversampling Method for Predicting Depression Remission in University Studentshttps://blog.schoergendorfer.com/post/a-lightweight-phenology-aware-yolov5-framework-for-tomato-growth-stage-detectionSecFeedde2026-09-01T14:33:13.370058+00:00A Lightweight Phenology-Aware YOLOv5 Framework for Tomato Growth Stage Detection in Resource-Constrained Bhutanese Greenhouse Environmentshttps://blog.schoergendorfer.com/post/tracing-generated-samples-to-training-data-clusters-in-flow-matching-modelsSecFeedde2026-09-01T14:33:13.347260+00:00Tracing Generated Samples to Training-Data Clusters in Flow-Matching Modelshttps://blog.schoergendorfer.com/post/selection-representation-and-execution-in-sparse-fourier-neural-operatorsSecFeedde2026-09-01T14:33:13.328094+00:00Selection, Representation, and Execution in Sparse Fourier Neural Operatorshttps://blog.schoergendorfer.com/post/multiclass-linear-perceptrons-with-multiplicative-marginsSecFeedde2026-09-01T14:33:13.291691+00:00Multiclass Linear Perceptrons with Multiplicative Marginshttps://blog.schoergendorfer.com/post/the-intervention-gap-in-latent-world-modelsSecFeedde2026-09-01T14:33:13.271610+00:00The Intervention Gap in Latent World Modelshttps://blog.schoergendorfer.com/post/robust-broad-learning-system-with-wave-loss-for-classification-under-data-uncertSecFeedde2026-09-01T14:33:13.247863+00:00Robust Broad Learning System with Wave Loss for Classification under Data Uncertaintyhttps://blog.schoergendorfer.com/post/diffusion-based-inverse-design-of-dielectric-resonator-metasurfaces-for-shapingSecFeedde2026-09-01T14:33:13.200427+00:00Diffusion-Based Inverse Design of Dielectric Resonator Metasurfaces for Shaping Smart Electromagnetic Environmentshttps://blog.schoergendorfer.com/post/joint-spatiotemporal-spectral-neural-operators-for-learning-pdes-on-irregular-doSecFeedde2026-09-01T14:33:13.176948+00:00Joint Spatiotemporal Spectral Neural Operators for Learning PDEs on Irregular Domainshttps://blog.schoergendorfer.com/post/sensitivity-constrained-neural-operators-for-data-efficient-forward-and-inverseSecFeedde2026-09-01T14:33:13.151508+00:00Sensitivity-Constrained Neural Operators for Data-Efficient Forward and Inverse Modeling of Partial Differential Equation Systemshttps://blog.schoergendorfer.com/post/structural-hierarchy-and-geometry-in-molecular-representation-learningSecFeedde2026-09-01T14:18:14.575987+00:00Structural Hierarchy and Geometry in Molecular Representation Learninghttps://blog.schoergendorfer.com/post/partially-linear-autoencoders-for-manifold-learning-and-dimensionality-reductionSecFeedde2026-09-01T14:18:14.534591+00:00Partially Linear Autoencoders for Manifold Learning and Dimensionality Reductionhttps://blog.schoergendorfer.com/post/uncertainty-driven-replay-memory-for-reinforcement-learningSecFeedde2026-09-01T14:18:14.506292+00:00Uncertainty-Driven Replay Memory for Reinforcement Learninghttps://blog.schoergendorfer.com/post/designing-for-the-next-click-bandits-for-real-time-page-layoutSecFeedde2026-09-01T14:18:14.478061+00:00Designing for the Next Click: Bandits for Real-Time Page Layouthttps://blog.schoergendorfer.com/post/structure-aware-neural-architecture-search-for-mixture-of-expertsSecFeedde2026-09-01T14:18:14.453797+00:00Structure Aware Neural Architecture Search for Mixture of Expertshttps://blog.schoergendorfer.com/post/pruneshift-a-framework-for-evaluating-decision-reliability-in-structured-pruningSecFeedde2026-09-01T14:18:14.429335+00:00PruneShift: A Framework for Evaluating Decision Reliability in Structured Pruninghttps://blog.schoergendorfer.com/post/eca-bls-an-efficient-complex-augmented-broad-learning-systemSecFeedde2026-09-01T14:18:14.407863+00:00ECA-BLS: An Efficient Complex-Augmented Broad Learning Systemhttps://blog.schoergendorfer.com/post/last-step-matters-early-uncertainty-cannot-predict-failure-in-long-horizon-agentSecFeedde2026-09-01T14:18:14.366082+00:00Last Step Matters: Early Uncertainty Cannot Predict Failure in Long-Horizon Agentshttps://blog.schoergendorfer.com/post/creation-begins-with-understanding-llms-as-strategy-designers-for-privacy-preserSecFeedde2026-09-01T14:18:14.344199+00:00Creation begins with understanding: LLMs as strategy designers for privacy-preserving tabular data synthesishttps://blog.schoergendorfer.com/post/reward-guided-fine-tuning-of-one-step-generative-models-via-wasserstein-gradientSecFeedde2026-09-01T14:18:14.299356+00:00Reward-guided Fine-Tuning of One-Step Generative Models via Wasserstein Gradient Flowhttps://blog.schoergendorfer.com/post/unsupervised-multi-scale-gromov-wasserstein-hypergraph-alignmentSecFeedde2026-09-01T14:18:14.277452+00:00Unsupervised Multi-Scale Gromov-Wasserstein Hypergraph Alignmenthttps://blog.schoergendorfer.com/post/adaptive-doubly-robust-off-policy-evaluation-for-ranking-policies-under-diverseSecFeedde2026-09-01T14:18:14.250674+00:00Adaptive Doubly Robust Off-Policy Evaluation for Ranking Policies under Diverse User Behaviorhttps://blog.schoergendorfer.com/post/on-the-resilience-of-text-to-video-diffusion-models-to-hardware-faultsSecFeedde2026-09-01T14:18:14.223079+00:00On the Resilience of Text-to-Video Diffusion Models to Hardware Faultshttps://blog.schoergendorfer.com/post/predicting-the-unpredictable-llm-powered-long-term-chaotic-time-series-forecastiSecFeedde2026-09-01T14:18:14.193584+00:00Predicting the Unpredictable: LLM-powered Long-term Chaotic Time Series Forecasting under Short-term Observationshttps://blog.schoergendorfer.com/post/event-triggered-control-and-online-learning-for-networked-systems-under-computatSecFeedde2026-09-01T14:18:14.168549+00:00Event-triggered Control and Online Learning for Networked Systems under Computational Delayshttps://blog.schoergendorfer.com/post/asynchronous-cooperative-online-learning-for-multi-robot-control-under-computatiSecFeedde2026-09-01T14:18:14.135761+00:00Asynchronous Cooperative Online Learning for Multi-Robot Control under Computational Delayshttps://blog.schoergendorfer.com/post/which-llm-for-which-work-budgeted-model-allocation-under-uncertain-evaluationSecFeedde2026-09-01T14:18:14.106240+00:00Which LLM for Which Work? Budgeted Model Allocation under Uncertain Evaluationhttps://blog.schoergendorfer.com/post/beacon-behavioral-and-semantic-enrichment-of-alphaearth-embeddings-through-tri-mSecFeedde2026-09-01T14:03:02.984062+00:00BEACON: Behavioral and Semantic Enrichment of AlphaEarth Embeddings through Tri-Modal Contrastive Learninghttps://blog.schoergendorfer.com/post/medcache-efficient-and-temporally-valid-memory-for-longitudinal-clinical-agentsSecFeedde2026-09-01T14:03:02.961428+00:00MedCache: Efficient and Temporally Valid Memory for Longitudinal Clinical Agentshttps://blog.schoergendorfer.com/post/adversarial-online-classification-with-a-previewSecFeedde2026-09-01T14:03:02.938068+00:00Adversarial Online Classification with a Previewhttps://blog.schoergendorfer.com/post/target-aware-state-adaptive-p-dirichlet-graph-neural-regression-for-non-invasiveSecFeedde2026-09-01T14:03:02.911594+00:00Target-Aware State-Adaptive $p$-Dirichlet Graph Neural Regression for Non-Invasive Body-Composition Estimationhttps://blog.schoergendorfer.com/post/learning-human-health-and-diseases-from-24-hour-wrist-movementSecFeedde2026-09-01T14:03:02.884563+00:00Learning Human Health and Diseases from 24-hour Wrist Movementhttps://blog.schoergendorfer.com/post/a-causal-model-for-locating-and-unlocking-sandbagging-in-model-organismsSecFeedde2026-09-01T14:03:02.858835+00:00A Causal Model for Locating and Unlocking Sandbagging in Model Organismshttps://blog.schoergendorfer.com/post/behavioral-latency-as-weak-event-time-supervision-for-eeg-reaction-time-decodingSecFeedde2026-09-01T14:03:02.834238+00:00Behavioral Latency as Weak Event-Time Supervision for EEG Reaction-Time Decodinghttps://blog.schoergendorfer.com/post/one-capability-or-many-testing-the-economic-validity-of-frontier-ai-evaluationSecFeedde2026-09-01T14:03:02.807533+00:00One Capability or Many? Testing the Economic Validity of Frontier AI Evaluationhttps://blog.schoergendorfer.com/post/where-induction-runs-out-description-length-difficulty-and-the-memorisation-gapSecFeedde2026-09-01T14:03:02.782635+00:00Where Induction Runs Out: Description-Length Difficulty and the Memorisation Gap in Integer-Sequence Benchmarkshttps://blog.schoergendorfer.com/post/unlearning-on-spatio-temporal-graphs-through-subgraph-virtual-edge-reconstructioSecFeedde2026-09-01T14:03:02.754315+00:00Unlearning on Spatio-Temporal Graphs through Subgraph Virtual Edge Reconstructionhttps://blog.schoergendorfer.com/post/information-based-calibration-of-uncertainty-quantification-in-product-of-expertSecFeedde2026-09-01T14:03:02.707202+00:00Information-Based Calibration of Uncertainty Quantification in Product-of-Experts Gaussian Process Modelshttps://blog.schoergendorfer.com/post/mel-coordinate-preserving-eeg-tokenization-for-fmri-translationSecFeedde2026-09-01T14:03:02.677604+00:00MEL: Coordinate-Preserving EEG Tokenization for fMRI Translationhttps://blog.schoergendorfer.com/post/rl-fat-reinforcement-learning-for-fair-adversarial-trainingSecFeedde2026-09-01T14:03:02.652356+00:00RL-FAT: Reinforcement Learning for Fair Adversarial Traininghttps://blog.schoergendorfer.com/post/a-spectral-identifiability-threshold-for-dissipative-rate-recovery-from-truncateSecFeedde2026-09-01T14:03:02.621776+00:00A Spectral Identifiability Threshold for Dissipative Rate Recovery from Truncated Liouvillian Spectrahttps://blog.schoergendorfer.com/post/halluprism-when-multimodal-uncertainty-should-diagnose-not-decideSecFeedde2026-09-01T14:03:02.593324+00:00HalluPrism: When Multimodal Uncertainty Should Diagnose, Not Decidehttps://blog.schoergendorfer.com/post/explainable-machine-learning-for-broadband-adoption-disparities-tract-level-predSecFeedde2026-09-01T14:03:02.568105+00:00Explainable Machine Learning for Broadband Adoption Disparities: Tract-Level Prediction and SHAP-Based Factor Profilinghttps://blog.schoergendorfer.com/post/temperature-adaptive-transformed-teacher-matchingSecFeedde2026-09-01T14:03:02.541266+00:00Temperature-Adaptive Transformed Teacher Matchinghttps://blog.schoergendorfer.com/post/pathguide-dynamic-classifier-free-guidance-via-on-policy-transport-alignmentSecFeedde2026-09-01T14:03:02.513353+00:00PathGuide: Dynamic Classifier-Free Guidance via On-Policy Transport Alignmenthttps://blog.schoergendorfer.com/post/selective-disclosure-of-hidden-directives-in-reasoning-models-behavioral-asymmetSecFeedde2026-09-01T13:47:44.349722+00:00Selective Disclosure of Hidden Directives in Reasoning Models: Behavioral Asymmetry and Steeringhttps://blog.schoergendorfer.com/post/pathbridger-subgoal-bridges-for-offline-goal-conditioned-reinforcement-learningSecFeedde2026-09-01T13:47:44.331114+00:00PathBridger: Subgoal Bridges for Offline Goal-Conditioned Reinforcement Learninghttps://blog.schoergendorfer.com/post/sparse-koopman-autoencoders-identify-local-dynamical-regimes-in-multibasin-systeSecFeedde2026-09-01T13:47:44.312568+00:00Sparse Koopman Autoencoders Identify Local Dynamical Regimes in Multibasin Systemshttps://blog.schoergendorfer.com/post/nve-a-separability-and-coverage-aware-internal-validation-metric-for-biclusterinSecFeedde2026-09-01T13:47:44.291414+00:00NVE: A Separability and Coverage-Aware Internal Validation Metric for Biclusteringhttps://blog.schoergendorfer.com/post/hybrid-semantic-context-enhanced-ensemble-learning-for-wind-power-ramp-event-forSecFeedde2026-09-01T13:47:44.270785+00:00Hybrid Semantic Context-Enhanced Ensemble Learning for Wind Power Ramp-Event Forecasting and Uncertainty-Aware Evaluationhttps://blog.schoergendorfer.com/post/effective-graph-and-rank-based-contextual-embeddings-for-textual-and-multimediaSecFeedde2026-09-01T13:47:44.249942+00:00Effective Graph and Rank-based Contextual Embeddings for Textual and Multimedia Datahttps://blog.schoergendorfer.com/post/context-aware-interpretable-representations-for-retrieval-and-graph-convolutionaSecFeedde2026-09-01T13:47:44.230232+00:00Context-Aware Interpretable Representations for Retrieval and Graph Convolutional Network Classificationhttps://blog.schoergendorfer.com/post/v2tatc-a-joint-voice-trajectory-embedding-framework-and-dataset-for-air-trafficSecFeedde2026-09-01T13:47:44.208115+00:00V2TATC: A Joint Voice-Trajectory Embedding Framework and Dataset for Air Traffic Controller Situational Awarenesshttps://blog.schoergendorfer.com/post/revisiting-the-provable-auditable-privacy-gap-of-dp-sgdSecFeedde2026-09-01T13:47:44.167806+00:00Revisiting the Provable-Auditable Privacy Gap of DP-SGDhttps://blog.schoergendorfer.com/post/rankshift-in-database-detection-and-explanation-of-categorical-shiftsSecFeedde2026-09-01T13:47:44.143033+00:00RankShift: In-Database Detection and Explanation of Categorical Shiftshttps://blog.schoergendorfer.com/post/semkv-semantic-mixed-precision-kv-cache-quantization-guided-by-the-quality-cliffSecFeedde2026-09-01T13:47:44.118578+00:00SemKV: Semantic Mixed-Precision KV Cache Quantization Guided by the Quality Cliff for Long-Context LLM Inferencehttps://blog.schoergendorfer.com/post/learning-theoretic-foundation-for-general-coded-computing-the-straggler-settingSecFeedde2026-09-01T13:47:44.097921+00:00Learning-Theoretic Foundation for General Coded Computing: The Straggler Settinghttps://blog.schoergendorfer.com/post/off-policy-evaluation-for-semantic-id-recommenders-does-the-models-own-code-hierSecFeedde2026-09-01T13:47:44.077275+00:00Off-Policy Evaluation for Semantic ID Recommenders: Does the Model's Own Code Hierarchy Help?https://blog.schoergendorfer.com/post/conservative-hybrid-graph-networks-for-process-systems-with-learned-routingSecFeedde2026-09-01T13:47:44.055819+00:00Conservative Hybrid Graph Networks for Process Systems with Learned Routinghttps://blog.schoergendorfer.com/post/unsupervised-latent-space-alignment-with-hyperspherical-geodesic-matchingSecFeedde2026-09-01T13:47:44.031327+00:00Unsupervised Latent Space Alignment with Hyperspherical Geodesic Matchinghttps://blog.schoergendorfer.com/post/err-sequential-entropy-resolution-for-efficient-and-decisive-llm-reasoningSecFeedde2026-09-01T13:47:44.004952+00:00ERR+: Sequential Entropy Resolution for Efficient and Decisive LLM Reasoninghttps://blog.schoergendorfer.com/post/deriving-scaling-laws-for-openeurollm-models-learning-rate-batch-size-and-loss-2SecFeedde2026-09-01T13:47:43.983439+00:00Deriving Scaling Laws for OpenEuroLLM Models: Learning Rate, Batch Size and Losshttps://blog.schoergendorfer.com/post/not-to-break-but-to-attest-adversarial-probes-for-privacy-preserving-llm-verific-2SecFeedde2026-09-01T13:47:43.959773+00:00Not to Break, but to Attest: Adversarial Probes for Privacy-Preserving LLM Verificationhttps://blog.schoergendorfer.com/post/a-survey-on-rubric-guided-reinforcement-learning-for-language-models-2SecFeedde2026-09-01T13:47:43.935092+00:00A Survey on Rubric-Guided Reinforcement Learning for Language Modelshttps://blog.schoergendorfer.com/post/xhotpotqa-a-benchmark-for-cross-lingual-knowledge-composition-in-multi-hop-quest-2SecFeedde2026-09-01T13:33:02.507655+00:00XHotpotQA: A Benchmark for Cross-Lingual Knowledge Composition in Multi-Hop Question Answeringhttps://blog.schoergendorfer.com/post/successive-capacity-growth-task-complexity-driven-width-and-depth-expansion-for-2SecFeedde2026-09-01T13:33:02.483411+00:00Successive Capacity Growth: Task-Complexity-Driven Width and Depth Expansion for Vision Transformer Encoders in JEPA World Modelshttps://blog.schoergendorfer.com/post/icon-decomposition-auditing-deep-neural-networks-with-multivariate-variance-baseSecFeedde2026-09-01T13:33:02.461688+00:00ICON Decomposition: Auditing Deep Neural Networks with Multivariate Variance-based Concept-level Explanationshttps://blog.schoergendorfer.com/post/a-statistical-audit-of-physical-ai-benchmark-redundancySecFeedde2026-09-01T13:33:02.440121+00:00A Statistical Audit of Physical AI Benchmark Redundancyhttps://blog.schoergendorfer.com/post/moplex-estimating-plackett-luce-mixture-models-for-multi-objective-alignmentSecFeedde2026-09-01T13:33:02.394907+00:00MoPLEx: Estimating Plackett-Luce Mixture Models for Multi-Objective Alignmenthttps://blog.schoergendorfer.com/post/demmo-longitudinal-and-cross-disease-modelling-of-digital-mobility-outcomes-via-2SecFeedde2026-09-01T13:33:02.371401+00:00DeMMO: Longitudinal and Cross-Disease Modelling of Digital Mobility Outcomes via Multi-Task Learninghttps://blog.schoergendorfer.com/post/demystifying-reinforcement-learning-post-training-of-language-models-2SecFeedde2026-09-01T13:33:02.346643+00:00Demystifying Reinforcement Learning Post-Training of Language Modelshttps://blog.schoergendorfer.com/post/fampwq-fisher-information-based-adaptive-mixed-precision-weight-quantization-for-2SecFeedde2026-09-01T13:33:02.324437+00:00FAMPWQ: Fisher Information-based Adaptive Mixed Precision Weight Quantization for Effective LLM Inferencehttps://blog.schoergendorfer.com/post/enhancing-bayesian-optimization-and-active-learning-through-kernel-diversity-2SecFeedde2026-09-01T13:33:02.298588+00:00Enhancing Bayesian Optimization and Active Learning Through Kernel Diversityhttps://blog.schoergendorfer.com/post/when-less-is-more-an-empirical-study-of-minimal-responses-in-counseling-dialogue-2SecFeedde2026-09-01T13:33:02.268997+00:00When Less Is More: An Empirical Study of Minimal Responses in Counseling Dialogues and the Behavior of LLMshttps://blog.schoergendorfer.com/post/the-shadow-price-of-intelligence-quality-degradation-in-llm-inference-as-a-suppl-2SecFeedde2026-09-01T13:33:02.247080+00:00The Shadow Price of Intelligence: Quality Degradation in LLM Inference as a Supply Chain Problemhttps://blog.schoergendorfer.com/post/lucaid-agentic-multimodal-ai-for-lung-cancer-precision-pathology-2SecFeedde2026-09-01T13:33:02.222329+00:00LUCAID: Agentic Multimodal AI for Lung Cancer Precision Pathologyhttps://blog.schoergendorfer.com/post/mycelial-search-a-graph-structured-metaheuristic-for-continuous-optimisation-2SecFeedde2026-09-01T13:33:02.193686+00:00Mycelial Search: A Graph-Structured Metaheuristic for Continuous Optimisationhttps://blog.schoergendorfer.com/post/literarybigfive-author-personalized-text-generation-in-a-unified-interpretable-s-3SecFeedde2026-09-01T13:33:02.164750+00:00LITERARYBIGFIVE: Author-Personalized Text Generation in a Unified Interpretable Spacehttps://blog.schoergendorfer.com/post/dele-w05-inferring-action-from-future-latent-state-for-robotic-manipulation-2SecFeedde2026-09-01T13:33:02.119712+00:00DELE-w0.5: Inferring Action from Future Latent State for Robotic Manipulationhttps://blog.schoergendorfer.com/post/reinforcement-learning-on-benign-facts-amplifies-leakage-of-memorized-private-da-3SecFeedde2026-09-01T13:33:02.088854+00:00Reinforcement Learning on Benign Facts Amplifies Leakage of Memorized Private Datahttps://blog.schoergendorfer.com/post/atom-learning-model-alm-how-a-real-classroom-got-tokenised-3SecFeedde2026-09-01T13:33:02.064647+00:00Atom Learning Model (ALM): how a real classroom got tokenisedhttps://blog.schoergendorfer.com/post/when-images-look-right-and-retrieve-wrong-coverage-guided-cross-scale-re-indexinSecFeedde2026-09-01T13:33:02.039538+00:00When Images Look Right and Retrieve Wrong: Coverage-Guided Cross-Scale Re-Indexing for Knowledge-Faithful Generative Perceptionhttps://blog.schoergendorfer.com/post/spade-self-play-in-adaptive-synthetic-executable-environments-3SecFeedde2026-09-01T13:17:44.437758+00:00SPADE: Self-Play in Adaptive Synthetic Executable Environmentshttps://blog.schoergendorfer.com/post/decomposing-wrong-consensus-agreement-in-llm-self-consistencySecFeedde2026-09-01T13:17:44.406513+00:00Decomposing Wrong-Consensus Agreement in LLM Self-Consistencyhttps://blog.schoergendorfer.com/post/modeling-the-structure-of-human-behavior-with-ai-prompt-vectorsSecFeedde2026-09-01T13:17:44.376520+00:00Modeling the Structure of Human Behavior with AI Prompt Vectorshttps://blog.schoergendorfer.com/post/low-rank-dynamics-effective-latent-carriers-for-counterfactual-rollout-in-learne-3SecFeedde2026-09-01T13:17:44.308030+00:00Low-Rank Dynamics-Effective Latent Carriers for Counterfactual Rollout in Learned World Modelshttps://blog.schoergendorfer.com/post/writing-style-similarity-reflects-academic-genealogy-2SecFeedde2026-09-01T13:17:44.288217+00:00Writing Style Similarity Reflects Academic Genealogyhttps://blog.schoergendorfer.com/post/reduced-matrix-multiplication-input-adaptive-matrix-product-reduction-for-llm-in-2SecFeedde2026-09-01T13:17:44.265298+00:00Reduced Matrix Multiplication: Input-Adaptive Matrix-Product Reduction for LLM Inferencehttps://blog.schoergendorfer.com/post/terminal-symmetry-as-a-carrier-of-asymmetric-process-knowledge-statewise-refinemSecFeedde2026-09-01T13:17:44.199903+00:00Terminal Symmetry as a Carrier of Asymmetric Process Knowledge: Statewise Refinement for Anytime Verified Constructionhttps://blog.schoergendorfer.com/post/surfacing-the-unsaid-cue-bench-for-affective-stance-in-chinese-discourse-3SecFeedde2026-09-01T13:17:44.164336+00:00Surfacing the Unsaid: CUE-Bench for Affective Stance in Chinese Discoursehttps://blog.schoergendorfer.com/post/auditing-chinese-web-scale-corpora-via-sampled-bpe-token-statistics-2SecFeedde2026-09-01T13:17:44.133772+00:00Auditing Chinese Web-scale Corpora via Sampled BPE Token Statisticshttps://blog.schoergendorfer.com/post/ouroboros-a-self-developing-frontier-coding-agent-with-reviewed-core-evolution-3SecFeedde2026-09-01T13:17:44.084679+00:00Ouroboros: A Self-Developing Frontier Coding Agent with Reviewed Core Evolutionhttps://blog.schoergendorfer.com/post/approximate-speculative-decoding-3SecFeedde2026-09-01T13:17:44.050860+00:00Approximate Speculative Decodinghttps://blog.schoergendorfer.com/post/emulate-or-estimate-the-divergent-strengths-of-base-and-post-trained-language-mo-2SecFeedde2026-09-01T13:17:44.026268+00:00Emulate or Estimate? The Divergent Strengths of Base and Post-Trained Language Models for Opinion Simulationhttps://blog.schoergendorfer.com/post/single-canonical-prompts-underestimate-llm-safetys-surface-form-sensitivity-2SecFeedde2026-09-01T13:17:44.000861+00:00Single Canonical Prompts Underestimate LLM Safety's Surface-Form Sensitivityhttps://blog.schoergendorfer.com/post/why-llms-give-in-conversational-factors-and-reasoning-behind-medical-sycophancy-2SecFeedde2026-09-01T13:17:43.969543+00:00Why LLMs Give In: Conversational Factors and Reasoning Behind Medical Sycophancyhttps://blog.schoergendorfer.com/post/serum-state-extraction-and-refinement-for-user-modeling-2SecFeedde2026-09-01T13:02:37.655038+00:00SERUM: State Extraction and Refinement for User Modelinghttps://blog.schoergendorfer.com/post/why-it-hurts-identifying-the-drivers-of-negative-thoughts-in-emotional-support-c-2SecFeedde2026-09-01T13:02:37.634005+00:00Why It Hurts: Identifying the Drivers of Negative Thoughts in Emotional Support Conversationshttps://blog.schoergendorfer.com/post/searching-for-robust-augmentations-to-improve-out-of-domain-generalization-in-de-2SecFeedde2026-09-01T13:02:37.610636+00:00Searching for Robust Augmentations to Improve Out-of-Domain Generalization in Dermoscopic Skin Cancer Classificationhttps://blog.schoergendorfer.com/post/learning-to-trace-seiberg-dualities-2SecFeedde2026-09-01T13:02:37.587301+00:00Learning to Trace Seiberg Dualitieshttps://blog.schoergendorfer.com/post/a2tta-anchored-and-agile-test-time-adaptation-for-evolving-traffic-sensor-networ-4SecFeedde2026-09-01T13:02:37.565218+00:00A2TTA: Anchored-and-Agile Test-Time Adaptation for Evolving Traffic Sensor Networkshttps://blog.schoergendorfer.com/post/syrup-enhancing-system-prompt-following-via-reward-guided-prediction-in-llm-deco-2SecFeedde2026-09-01T13:02:37.536782+00:00SyRuP: Enhancing System-Prompt Following via Reward-Guided Prediction in LLM Decodinghttps://blog.schoergendorfer.com/post/action-from-adjacent-set-in-physical-space-outperforms-the-best-prediction-in-wo-2SecFeedde2026-09-01T13:02:37.512505+00:00Action from Adjacent Set in Physical Space Outperforms the Best Prediction in World Modelshttps://blog.schoergendorfer.com/post/error-certificates-for-kv-cache-eviction-via-randomized-design-3SecFeedde2026-09-01T13:02:37.491385+00:00Error Certificates for KV-Cache Eviction via Randomized Designhttps://blog.schoergendorfer.com/post/off-context-grpo-learning-to-reason-on-hard-problems-using-privileged-informatio-2SecFeedde2026-09-01T13:02:37.461789+00:00Off-Context GRPO: Learning to Reason on Hard Problems using Privileged Informationhttps://blog.schoergendorfer.com/post/reviv-reconstructing-the-viewer-and-the-view-in-4d-from-monocular-egocentric-vid-2SecFeedde2026-09-01T13:02:37.435092+00:00ReViV: Reconstructing the Viewer and the View in 4D from Monocular Egocentric Videohttps://blog.schoergendorfer.com/post/governing-well-in-the-algorithmic-age-the-foundations-of-digital-statecraft-3SecFeedde2026-09-01T13:02:37.412741+00:00Governing Well in the Algorithmic Age: The Foundations of Digital Statecrafthttps://blog.schoergendorfer.com/post/salt-salience-aware-lexical-trie-for-long-context-compression-2SecFeedde2026-09-01T13:02:37.389579+00:00SALT: Salience-Aware Lexical Trie for Long-Context Compressionhttps://blog.schoergendorfer.com/post/memora-embodied-action-memory-from-egocentric-videos-for-reasoning-and-planning-2SecFeedde2026-09-01T13:02:37.359456+00:00MEMORA: Embodied Action Memory from Egocentric Videos for Reasoning and Planninghttps://blog.schoergendorfer.com/post/decoupled-structure-feature-alignment-via-alternating-optimization-for-graph-leaSecFeedde2026-09-01T13:02:37.335430+00:00Decoupled Structure-Feature Alignment via Alternating Optimization for Graph Learninghttps://blog.schoergendorfer.com/post/prism-edit-one-vector-for-all-temporal-answers-2SecFeedde2026-09-01T13:02:37.312219+00:00PRISM Edit: One Vector for All Temporal Answershttps://blog.schoergendorfer.com/post/synthave-scalable-synthetic-labeling-for-e-commerce-with-llm-arena-validation-2SecFeedde2026-09-01T13:02:37.289366+00:00SynthAVE: Scalable Synthetic Labeling for E-Commerce with LLM-Arena Validationhttps://blog.schoergendorfer.com/post/a-gold-standard-study-of-what-makes-a-lightweight-game-playing-agent-strong-2SecFeedde2026-09-01T13:02:37.263247+00:00A Gold-Standard Study of What Makes a Lightweight Game-Playing Agent Stronghttps://blog.schoergendorfer.com/post/a-few-teacher-steps-go-a-long-way-cost-efficient-on-policy-data-augmentation-for-2SecFeedde2026-09-01T13:02:37.238910+00:00A Few Teacher Steps Go a Long Way: Cost-Efficient On-Policy Data Augmentation for Agent Post-Traininghttps://blog.schoergendorfer.com/post/criterion-conditional-in-context-learning-evaluating-criterion-shift-adaptation-2SecFeedde2026-09-01T13:02:37.212040+00:00Criterion-Conditional In-Context Learning: Evaluating Criterion-Shift Adaptation in Vision-Language Modelshttps://blog.schoergendorfer.com/post/cultural-bias-without-a-cultural-selfa-disassociation-study-of-llms-persona-andSecFeedde2026-09-01T13:02:37.188136+00:00Cultural Bias Without a Cultural Self:A Disassociation Study of LLM's Persona and Biashttps://blog.schoergendorfer.com/post/scene-conditioned-pinn-gnn-for-multipath-rf-maps-cross-scene-generation-and-in-sSecFeedde2026-09-01T12:48:17.313603+00:00Scene-Conditioned PINN-GNN for Multipath RF Maps: Cross-Scene Generation and In-Scene Completionhttps://blog.schoergendorfer.com/post/memory-native-non-terrestrial-networks-for-embodied-intelligenceSecFeedde2026-09-01T12:48:17.292589+00:00Memory-Native Non-Terrestrial Networks for Embodied Intelligencehttps://blog.schoergendorfer.com/post/self-organized-conformal-prediction-reducing-regional-coverage-gaps-with-unsuperSecFeedde2026-09-01T12:48:17.272019+00:00Self-Organized Conformal Prediction: Reducing Regional Coverage Gaps with Unsupervised Group Discoveryhttps://blog.schoergendorfer.com/post/predicting-metastatic-risk-from-primary-cancer-tissue-architecture-via-distanceSecFeedde2026-09-01T12:48:17.246359+00:00Predicting Metastatic Risk from Primary Cancer Tissue Architecture via Distance-Aware Spatial Modelinghttps://blog.schoergendorfer.com/post/multihashformer-hash-based-generative-language-modelsSecFeedde2026-09-01T12:48:17.222541+00:00MultiHashFormer: Hash-based Generative Language Modelshttps://blog.schoergendorfer.com/post/dmv-bench-diagnosing-long-horizon-multimodal-agents-visual-memory-with-incidentaSecFeedde2026-09-01T12:48:17.193640+00:00DMV-Bench: Diagnosing Long-Horizon Multimodal Agents' Visual Memory with Incidental Cue Injectionhttps://blog.schoergendorfer.com/post/data-intelligence-agents-interpreting-modeling-and-querying-enterprise-data-viaSecFeedde2026-09-01T12:48:17.152217+00:00Data Intelligence Agents: Interpreting, Modeling, and Querying Enterprise Data via Autonomous Coding Agentshttps://blog.schoergendorfer.com/post/are-llms-ready-to-assist-physicians-physassistbench-for-interactive-doctor-patie-2SecFeedde2026-09-01T12:48:17.126851+00:00Are LLMs Ready to Assist Physicians? PhysAssistBench for Interactive Doctor-Patient-EHR Assistancehttps://blog.schoergendorfer.com/post/apt-atomic-physical-transitions-for-causal-video-language-understandingSecFeedde2026-09-01T12:48:17.106312+00:00APT: Atomic Physical Transitions for Causal Video-Language Understandinghttps://blog.schoergendorfer.com/post/magicsim-a-unified-infrastructure-for-executable-embodied-interactionSecFeedde2026-09-01T12:48:17.085492+00:00MagicSim: A Unified Infrastructure for Executable Embodied Interactionhttps://blog.schoergendorfer.com/post/care-context-aware-ranking-evolution-with-executable-scoring-programs-for-budget-2SecFeedde2026-09-01T12:48:17.029640+00:00CARE: Context-Aware Ranking Evolution with Executable Scoring Programs for Budgeted Reaction Optimizationhttps://blog.schoergendorfer.com/post/trace-trajectory-routed-causal-memory-for-delayed-evidence-visuomotor-imitationSecFeedde2026-09-01T12:48:17.003042+00:00TRACE: Trajectory-Routed Causal Memory for Delayed-Evidence Visuomotor Imitationhttps://blog.schoergendorfer.com/post/biodivergence-a-benchmark-and-evaluation-framework-for-hidden-contextual-contradSecFeedde2026-09-01T12:48:16.961509+00:00BioDivergence: A Benchmark and Evaluation Framework for Hidden Contextual Contradictions in Biomedical Abstractshttps://blog.schoergendorfer.com/post/attention-discounted-adaptive-sampler-for-masked-diffusion-language-models-2SecFeedde2026-09-01T12:48:16.936947+00:00Attention-Discounted Adaptive Sampler for Masked Diffusion Language Modelshttps://blog.schoergendorfer.com/post/able-representing-and-mapping-llms-via-attribution-based-large-model-embeddingSecFeedde2026-09-01T12:32:57.836337+00:00ABLE: Representing and Mapping LLMs via Attribution-Based Large-model Embeddinghttps://blog.schoergendorfer.com/post/twelve-quick-tips-for-designing-ai-driven-hpc-workflowsSecFeedde2026-09-01T12:32:57.814134+00:00Twelve quick tips for designing AI-driven HPC workflowshttps://blog.schoergendorfer.com/post/itp-stdp-a-hardware-efficient-intrinsic-timing-power-of-two-synaptic-learning-enSecFeedde2026-09-01T12:32:57.793493+00:00ITP-STDP: A Hardware-Efficient Intrinsic-Timing Power-of-Two Synaptic Learning Engine for On-Chip SNNshttps://blog.schoergendorfer.com/post/polaris-guiding-small-models-to-write-long-storiesSecFeedde2026-09-01T12:32:57.773339+00:00POLARIS: Guiding Small Models to Write Long Storieshttps://blog.schoergendorfer.com/post/qubric-co-designing-queries-and-rubrics-for-rl-beyond-verifiable-rewardsSecFeedde2026-09-01T12:32:57.753186+00:00QUBRIC: Co-Designing Queries and Rubrics for RL Beyond Verifiable Rewardshttps://blog.schoergendorfer.com/post/agentic-chain-of-thought-steering-for-efficient-and-controllable-llm-reasoningSecFeedde2026-09-01T12:32:57.731870+00:00Agentic Chain-of-Thought Steering for Efficient and Controllable LLM Reasoninghttps://blog.schoergendorfer.com/post/the-unsampled-truth-quantifying-prompt-artifacts-in-lm-psychometricsSecFeedde2026-09-01T12:32:57.711478+00:00The Unsampled Truth: Quantifying Prompt Artifacts in LM Psychometricshttps://blog.schoergendorfer.com/post/openagenet-oan-white-paper-open-infrastructure-for-trusted-agent-interconnectionSecFeedde2026-09-01T12:32:57.671685+00:00OpenAgenet / OAN White Paper: Open Infrastructure for Trusted Agent Interconnectionhttps://blog.schoergendorfer.com/post/fsa-grpo-teaching-auditory-llms-to-use-few-shot-demonstrationsSecFeedde2026-09-01T12:32:57.633129+00:00FSA-GRPO: Teaching Auditory LLMs to Use Few-Shot Demonstrationshttps://blog.schoergendorfer.com/post/dash-dual-branch-score-distillation-for-guidance-calibrated-compact-diffusion-moSecFeedde2026-09-01T12:32:57.613094+00:00DASH: Dual-Branch Score Distillation for Guidance-Calibrated Compact Diffusion Modelshttps://blog.schoergendorfer.com/post/the-value-of-spike-timing-a-leakage-resistant-benchmark-of-snn-design-choices-foSecFeedde2026-09-01T12:32:57.593371+00:00The Value of Spike Timing: A Leakage-Resistant Benchmark of SNN Design Choices for Network Intrusion Detectionhttps://blog.schoergendorfer.com/post/linguistics-aware-non-distortionary-llm-watermarkingSecFeedde2026-09-01T12:32:57.573870+00:00Linguistics-Aware Non-Distortionary LLM Watermarkinghttps://blog.schoergendorfer.com/post/skill-or-skip-learning-selective-skill-invocation-in-agentic-tasks-via-dual-granSecFeedde2026-09-01T12:32:57.551195+00:00Skill or Skip? Learning Selective Skill Invocation in Agentic Tasks via Dual-Granularity Preference Learninghttps://blog.schoergendorfer.com/post/vision-language-models-suppress-female-representations-under-ambiguous-inputSecFeedde2026-09-01T12:32:57.526985+00:00Vision-Language Models Suppress Female Representations Under Ambiguous Inputhttps://blog.schoergendorfer.com/post/exploring-autonomous-agentic-data-engineering-for-model-specializationSecFeedde2026-09-01T12:32:57.489951+00:00Exploring Autonomous Agentic Data Engineering for Model Specializationhttps://blog.schoergendorfer.com/post/lara-layer-wise-representation-analysis-for-detecting-data-contamination-in-rl-pSecFeedde2026-09-01T12:32:57.470721+00:00LaRA: Layer-wise Representation Analysis for Detecting Data Contamination in RL Post-Traininghttps://blog.schoergendorfer.com/post/how-coding-agents-fail-their-users-a-large-scale-analysis-of-developer-agent-misSecFeedde2026-09-01T12:32:57.448310+00:00How Coding Agents Fail Their Users: A Large-Scale Analysis of Developer-Agent Misalignment in 20,574 Real-World Sessionshttps://blog.schoergendorfer.com/post/relevance-as-a-vulnerability-how-web-retrieval-degrades-safety-alignment-in-llmSecFeedde2026-09-01T12:18:05.114073+00:00Relevance as a Vulnerability: How Web Retrieval Degrades Safety Alignment in LLM Agentshttps://blog.schoergendorfer.com/post/oisd-on-policy-internal-self-distillation-of-language-modelsSecFeedde2026-09-01T12:18:05.091220+00:00OISD: On-Policy Internal Self-Distillation of Language Modelshttps://blog.schoergendorfer.com/post/skill-conditioned-gated-self-distillation-for-llm-reasoning-2SecFeedde2026-09-01T12:18:05.069880+00:00Skill-Conditioned Gated Self-Distillation for LLM Reasoninghttps://blog.schoergendorfer.com/post/reverse-probing-supervised-token-level-uncertainty-quantification-for-large-langSecFeedde2026-09-01T12:18:05.047725+00:00Reverse Probing: Supervised Token-level Uncertainty Quantification for Large Language Models in Clinical Texthttps://blog.schoergendorfer.com/post/benger-benchmarking-llm-systems-on-subsumption-based-legal-reasoning-in-german-lSecFeedde2026-09-01T12:18:05.025189+00:00BenGER: Benchmarking LLM Systems on Subsumption-Based Legal Reasoning in German Lawhttps://blog.schoergendorfer.com/post/extracting-small-translation-specialists-from-llms-by-aggressively-pruning-experSecFeedde2026-09-01T12:18:05.004706+00:00Extracting Small Translation Specialists from LLMs by Aggressively Pruning Expertshttps://blog.schoergendorfer.com/post/kvoicebench-kopenaudiobench-and-kmmau-agent-driven-korean-speech-benchmarks-forSecFeedde2026-09-01T12:18:04.985234+00:00KVoiceBench, KOpenAudioBench, and KMMAU: Agent-Driven Korean Speech Benchmarks for Evaluating SpeechLMshttps://blog.schoergendorfer.com/post/integrated-and-cross-architecture-interpretation-of-llm-reasoningSecFeedde2026-09-01T12:18:04.963762+00:00Integrated and Cross-Architecture Interpretation of LLM Reasoninghttps://blog.schoergendorfer.com/post/quack-questioning-understanding-and-auditing-communicated-knowledge-in-multimodaSecFeedde2026-09-01T12:18:04.882634+00:00QUACK: Questioning, Understanding, and Auditing Communicated Knowledge in Multimodal Social Deduction Agentshttps://blog.schoergendorfer.com/post/when-the-strongest-teacher-is-not-the-best-teacher-student-centric-answer-selectSecFeedde2026-09-01T12:18:04.861124+00:00When the Strongest Teacher Is Not the Best Teacher: Student-Centric Answer Selectionhttps://blog.schoergendorfer.com/post/measuring-the-depth-of-llm-unlearning-via-activation-patchingSecFeedde2026-09-01T12:18:04.841086+00:00Measuring the Depth of LLM Unlearning via Activation Patchinghttps://blog.schoergendorfer.com/post/making-the-discrete-continuous-synthetic-raw-augmentations-for-fine-grained-evalSecFeedde2026-09-01T12:18:04.819140+00:00Making the Discrete Continuous: Synthetic RAW Augmentations for Fine-Grained Evaluation of Person Detection Performance in Low Lighthttps://blog.schoergendorfer.com/post/adversarial-trust-poisoning-in-vehicular-collaborative-perceptionSecFeedde2026-09-01T12:18:04.791298+00:00Adversarial Trust Poisoning in Vehicular Collaborative Perceptionhttps://blog.schoergendorfer.com/post/prefix-adaptive-block-diffusion-for-efficient-document-recognitionSecFeedde2026-09-01T12:18:04.744559+00:00Prefix-Adaptive Block Diffusion for Efficient Document Recognitionhttps://blog.schoergendorfer.com/post/drop-the-act-probe-filtered-rl-for-faithful-chain-of-thought-reasoningSecFeedde2026-09-01T12:18:04.717543+00:00Drop the Act: Probe-Filtered RL for Faithful Chain-of-Thought Reasoninghttps://blog.schoergendorfer.com/post/large-language-models-for-agentic-netops-and-aiops-architectures-evaluation-andSecFeedde2026-09-01T12:18:04.689069+00:00Large Language Models for Agentic NetOps and AIOps: Architectures, Evaluation, and Safetyhttps://blog.schoergendorfer.com/post/concepts-whisper-spectral-anti-concentration-and-the-dual-geometry-of-transformeSecFeedde2026-09-01T12:02:53.214078+00:00Concepts Whisper: Spectral Anti-Concentration and the Dual Geometry of Transformer Representationshttps://blog.schoergendorfer.com/post/airfm-dda-air-interface-foundation-model-in-the-delay-doppler-angle-domain-for-aSecFeedde2026-09-01T12:02:53.191075+00:00AirFM-DDA: Air-Interface Foundation Model in the Delay-Doppler-Angle Domain for AI-Native 6Ghttps://blog.schoergendorfer.com/post/star-fusion-a-multi-modal-transformer-architecture-for-discrete-celestial-orientSecFeedde2026-09-01T12:02:53.168553+00:00Star-Fusion: A Multi-modal Transformer Architecture for Discrete Celestial Orientation via Spherical Topologyhttps://blog.schoergendorfer.com/post/secret-stealing-attacks-on-local-llm-fine-tuning-through-supply-chain-model-codeSecFeedde2026-09-01T12:02:53.146539+00:00Secret Stealing Attacks on Local LLM Fine-Tuning through Supply-Chain Model Code Backdoorshttps://blog.schoergendorfer.com/post/cf-vla-efficient-coarse-to-fine-action-generation-for-vision-language-action-polSecFeedde2026-09-01T12:02:53.123303+00:00CF-VLA: Efficient Coarse-to-Fine Action Generation for Vision-Language-Action Policieshttps://blog.schoergendorfer.com/post/inverting-foundation-models-of-brain-function-with-simulation-based-inferenceSecFeedde2026-09-01T12:02:53.104497+00:00Inverting Foundation Models of Brain Function with Simulation-Based Inferencehttps://blog.schoergendorfer.com/post/when-chain-of-thought-fails-the-solution-hides-in-the-hidden-statesSecFeedde2026-09-01T12:02:53.085925+00:00When Chain-of-Thought Fails, the Solution Hides in the Hidden Stateshttps://blog.schoergendorfer.com/post/sketchvlm-vision-language-models-can-annotate-images-to-explain-thoughts-and-guiSecFeedde2026-09-01T12:02:53.066820+00:00SketchVLM: Vision language models can annotate images to explain thoughts and guide usershttps://blog.schoergendorfer.com/post/breaking-mcp-with-function-hijacking-attacks-novel-threats-for-function-callingSecFeedde2026-09-01T12:02:53.049104+00:00Breaking MCP with Function Hijacking Attacks: Novel Threats for Function Calling and Agentic Modelshttps://blog.schoergendorfer.com/post/koala-bench-evaluating-large-audio-language-models-on-korean-speech-understandinSecFeedde2026-09-01T12:02:53.030612+00:00KoALa-Bench: Evaluating Large Audio Language Models on Korean Speech Understanding and Faithfulnesshttps://blog.schoergendorfer.com/post/perturbation-sensitivity-of-maximum-likelihood-pairwise-ranking-in-computationalSecFeedde2026-09-01T12:02:53.011019+00:00Perturbation Sensitivity of Maximum-Likelihood Pairwise Ranking in Computational Decision Systemshttps://blog.schoergendorfer.com/post/do-we-still-need-humans-in-the-loop-human-vs-llm-annotation-in-active-learning-fSecFeedde2026-09-01T12:02:52.988697+00:00Do We Still Need Humans in the Loop? Human vs. LLM Annotation in Active Learning for TikTok Hate Speech Detectionhttps://blog.schoergendorfer.com/post/unveiling-language-routing-isolation-in-multilingual-moe-models-for-interpretablSecFeedde2026-09-01T12:02:52.960708+00:00Unveiling Language Routing Isolation in Multilingual MoE Models for Interpretable Subnetwork Adaptationhttps://blog.schoergendorfer.com/post/interactive-clarification-for-cloud-infrastructure-as-code-synthesisSecFeedde2026-09-01T12:02:52.937734+00:00Interactive Clarification for Cloud Infrastructure-as-Code Synthesishttps://blog.schoergendorfer.com/post/where-does-robustness-live-neuron-guided-adaptation-for-retrieval-augmented-langSecFeedde2026-09-01T12:02:52.913924+00:00Where Does Robustness Live? Neuron-Guided Adaptation for Retrieval-Augmented Language Modelshttps://blog.schoergendorfer.com/post/robust-multi-agent-reinforcement-learning-for-small-uas-separation-assurance-und-2SecFeedde2026-09-01T12:02:52.887693+00:00Robust Multi-Agent Reinforcement Learning for Small UAS Separation Assurance under GPS Degradation and Spoofinghttps://blog.schoergendorfer.com/post/m2k-making-the-model-kernel-interface-explicit-for-reliable-cuda-kernel-verificaSecFeedde2026-09-01T11:48:18.088794+00:00M2K: Making the Model-Kernel Interface Explicit for Reliable CUDA Kernel Verificationhttps://blog.schoergendorfer.com/post/mixture-greedy-for-online-generative-model-selection-is-ucb-necessary-in-diversiSecFeedde2026-09-01T11:48:18.035355+00:00Mixture-Greedy for Online Generative Model Selection: Is UCB Necessary in Diversity-Aware Multi-Armed Bandits?https://blog.schoergendorfer.com/post/phygile-physics-prefix-guided-motion-generation-for-agile-general-humanoid-motioSecFeedde2026-09-01T11:48:18.015744+00:00PhyGile: Physics-Prefix Guided Motion Generation for Agile General Humanoid Motion Trackinghttps://blog.schoergendorfer.com/post/do-large-language-models-possess-a-theory-of-mind-a-comparative-evaluation-usingSecFeedde2026-09-01T11:48:17.994254+00:00Do Large Language Models Possess a Theory of Mind? A Comparative Evaluation Using the Strange Stories Paradigmhttps://blog.schoergendorfer.com/post/interpretable-predictability-based-ai-text-detection-a-replication-studySecFeedde2026-09-01T11:48:17.972761+00:00Interpretable Predictability-Based AI Text Detection: A Replication Studyhttps://blog.schoergendorfer.com/post/the-landscape-of-generative-ai-in-information-systems-a-synthesis-of-secondary-rSecFeedde2026-09-01T11:48:17.915975+00:00The Landscape of Generative AI in Information Systems: A Synthesis of Secondary Reviews and Research Agendashttps://blog.schoergendorfer.com/post/erecu-pseudo-label-evolution-fusion-and-refinement-with-multi-cue-learning-for-uSecFeedde2026-09-01T11:48:17.896488+00:00EReCu: Pseudo-label Evolution Fusion and Refinement with Multi-Cue Learning for Unsupervised Camouflage Detectionhttps://blog.schoergendorfer.com/post/grammar-of-the-wave-towards-explainable-multivariate-time-series-event-detectionSecFeedde2026-09-01T11:48:17.876089+00:00Grammar of the Wave: Towards Explainable Multivariate Time Series Event Detection via Neuro-Symbolic VLM Agentshttps://blog.schoergendorfer.com/post/efficient-cross-view-localization-in-6g-space-air-ground-integrated-network-2SecFeedde2026-09-01T11:48:17.854415+00:00Efficient Cross-View Localization in 6G Space-Air-Ground Integrated Networkhttps://blog.schoergendorfer.com/post/personalized-group-relative-policy-optimization-for-heterogenous-preference-aligSecFeedde2026-09-01T11:48:17.812129+00:00Personalized Group Relative Policy Optimization for Heterogenous Preference Alignmenthttps://blog.schoergendorfer.com/post/scale-plan-scalable-language-enabled-task-planning-for-heterogeneous-multi-robotSecFeedde2026-09-01T11:48:17.794330+00:00Scale-Plan: Scalable Language-Enabled Task Planning for Heterogeneous Multi-Robot Teamshttps://blog.schoergendorfer.com/post/mind-unified-inquiry-and-diagnosis-rl-with-criteria-grounded-clinical-supports-fSecFeedde2026-09-01T11:48:17.769136+00:00MIND: Unified Inquiry and Diagnosis RL with Criteria Grounded Clinical Supports for Psychiatric Consultationhttps://blog.schoergendorfer.com/post/phys4d-fine-grained-physics-consistent-4d-modeling-from-video-diffusionSecFeedde2026-09-01T11:48:17.747839+00:00Phys4D: Fine-Grained Physics-Consistent 4D Modeling from Video Diffusionhttps://blog.schoergendorfer.com/post/toward-generalizable-deep-learning-based-peatland-fire-detection-via-walsh-hadamSecFeedde2026-09-01T11:48:17.726731+00:00Toward Generalizable Deep Learning Based Peatland Fire Detection via Walsh Hadamard Transform and Domain Adaptationhttps://blog.schoergendorfer.com/post/designascode-bridging-structural-editability-and-visual-fidelity-in-graphic-desiSecFeedde2026-09-01T11:33:00.839620+00:00DesignAsCode: Bridging Structural Editability and Visual Fidelity in Graphic Design Generationhttps://blog.schoergendorfer.com/post/reverse-n-wise-output-oriented-testing-for-aiml-and-quantum-computing-systemsSecFeedde2026-09-01T11:33:00.802585+00:00Reverse N-Wise Output-Oriented Testing for AI/ML and Quantum Computing Systemshttps://blog.schoergendorfer.com/post/small-updates-big-doubts-does-parameter-efficient-fine-tuning-enhance-hallucinatSecFeedde2026-09-01T11:33:00.764924+00:00Small Updates, Big Doubts: Does Parameter-Efficient Fine-tuning Enhance Hallucination Detection ?https://blog.schoergendorfer.com/post/artifact-reduction-in-undersampled-3d-cone-beam-cts-using-a-hybrid-2d-3d-cnn-fraSecFeedde2026-09-01T11:33:00.742427+00:00Artifact Reduction in Undersampled 3D Cone-Beam CTs using a Hybrid 2D-3D CNN Frameworkhttps://blog.schoergendorfer.com/post/robustness-of-vision-language-models-against-split-image-harmful-input-attacksSecFeedde2026-09-01T11:33:00.721915+00:00Robustness of Vision Language Models Against Split-Image Harmful Input Attackshttps://blog.schoergendorfer.com/post/zero-shot-generalizable-graph-anomaly-detection-with-mixture-of-riemannian-experSecFeedde2026-09-01T11:33:00.697703+00:00Zero-shot Generalizable Graph Anomaly Detection with Mixture of Riemannian Expertshttps://blog.schoergendorfer.com/post/toward-scalable-audio-description-quality-control-a-workflow-for-evaluating-humaSecFeedde2026-09-01T11:33:00.633768+00:00Toward Scalable Audio Description Quality Control: A Workflow for Evaluating Human and VLM Ratershttps://blog.schoergendorfer.com/post/securing-time-integrity-in-energy-iot-against-clock-drift-and-y2k38-failuresSecFeedde2026-09-01T11:33:00.611518+00:00Securing Time Integrity in Energy IoT Against Clock Drift and Y2K38 Failureshttps://blog.schoergendorfer.com/post/beyond-factual-qa-mentorship-oriented-question-answering-over-long-form-multilin-2SecFeedde2026-09-01T11:33:00.570438+00:00Beyond Factual QA: Mentorship-Oriented Question Answering over Long-Form Multilingual Contenthttps://blog.schoergendorfer.com/post/standardizing-longitudinal-radiology-report-evaluation-via-large-language-modelSecFeedde2026-09-01T11:33:00.548308+00:00Standardizing Longitudinal Radiology Report Evaluation via Large Language Model Annotationhttps://blog.schoergendorfer.com/post/mechanism-shift-during-post-training-from-autoregressive-to-masked-diffusion-lanSecFeedde2026-09-01T11:33:00.524418+00:00Mechanism Shift During Post-training from Autoregressive to Masked Diffusion Language Modelshttps://blog.schoergendorfer.com/post/poci-diff-3d-layout-guided-diffusion-for-controllable-synthetic-surveillance-datSecFeedde2026-09-01T11:33:00.503168+00:00POCI-Diff: 3D-Layout Guided Diffusion for Controllable Synthetic Surveillance Data Generationhttps://blog.schoergendorfer.com/post/core-t-coherent-retrieval-of-tables-for-text-to-sqlSecFeedde2026-09-01T11:33:00.478497+00:00CORE-T: COherent REtrieval of Tables for Text-to-SQLhttps://blog.schoergendorfer.com/post/agentrim-tool-risk-mitigation-for-agentic-aiSecFeedde2026-09-01T11:33:00.453170+00:00AgenTRIM: Tool Risk Mitigation for Agentic AIhttps://blog.schoergendorfer.com/post/to-retrieve-or-to-think-cross-boundary-context-evolution-for-multi-hop-complex-rSecFeedde2026-09-01T11:18:07.042381+00:00To Retrieve or To Think? Cross-Boundary Context Evolution for Multi-hop Complex Reasoninghttps://blog.schoergendorfer.com/post/triggering-chain-of-thought-via-latent-feature-interventions-in-large-language-mSecFeedde2026-09-01T11:18:07.017839+00:00Triggering Chain-of-Thought via Latent Feature Interventions in Large Language Modelshttps://blog.schoergendorfer.com/post/kinship-data-benchmark-for-multi-hop-reasoningSecFeedde2026-09-01T11:18:06.994225+00:00Kinship Data Benchmark for Multi-hop Reasoninghttps://blog.schoergendorfer.com/post/do-language-models-reason-across-languagesSecFeedde2026-09-01T11:18:06.967718+00:00Do Language Models Reason Across Languages?https://blog.schoergendorfer.com/post/adafuse-adaptive-ensemble-decoding-with-test-time-scaling-for-llmsSecFeedde2026-09-01T11:18:06.924107+00:00AdaFuse: Adaptive Ensemble Decoding with Test-Time Scaling for LLMshttps://blog.schoergendorfer.com/post/epiqal-benchmarking-large-language-models-in-epidemiological-question-answeringSecFeedde2026-09-01T11:18:06.903064+00:00EpiQAL: Benchmarking Large Language Models in Epidemiological Question Answering and Reasoninghttps://blog.schoergendorfer.com/post/dip-dynamic-in-context-planner-for-diffusion-language-modelsSecFeedde2026-09-01T11:18:06.880599+00:00DIP: Dynamic In-Context Planner For Diffusion Language Modelshttps://blog.schoergendorfer.com/post/entropy-aware-token-rejection-for-improving-speculative-decodingSecFeedde2026-09-01T11:18:06.856813+00:00Entropy-Aware Token Rejection for Improving Speculative Decodinghttps://blog.schoergendorfer.com/post/scoring-reasoning-and-selecting-the-best-ensembling-large-language-models-via-aSecFeedde2026-09-01T11:18:06.833736+00:00Scoring, Reasoning, and Selecting the Best! Ensembling Large Language Models via a Peer-Review Processhttps://blog.schoergendorfer.com/post/kascade-a-practical-sparse-attention-method-for-long-context-llm-inferenceSecFeedde2026-09-01T11:18:06.791197+00:00Kascade: A Practical Sparse Attention Method for Long-Context LLM Inferencehttps://blog.schoergendorfer.com/post/sgm-safety-glasses-for-multimodal-large-language-models-via-neuron-level-detoxifSecFeedde2026-09-01T11:18:06.765113+00:00SGM: Safety Glasses for Multimodal Large Language Models via Neuron-Level Detoxificationhttps://blog.schoergendorfer.com/post/latency-response-theory-model-evaluating-large-language-models-via-response-accu-2SecFeedde2026-09-01T11:18:06.741937+00:00Latency-Response Theory Model: Evaluating Large Language Models via Response Accuracy and Chain-of-Thought Lengthhttps://blog.schoergendorfer.com/post/better-world-models-can-lead-to-better-post-training-performanceSecFeedde2026-09-01T11:18:06.715090+00:00Better World Models Can Lead to Better Post-Training Performancehttps://blog.schoergendorfer.com/post/scaleprm-training-process-reward-models-by-scaling-verification-compute-withoutSecFeedde2026-09-01T11:18:06.693576+00:00ScalePRM: Training Process Reward Models by Scaling Verification Compute Without Ground Truthhttps://blog.schoergendorfer.com/post/cofirec-coarse-to-fine-tokenization-for-generative-recommendationSecFeedde2026-09-01T11:18:06.670587+00:00CoFiRec: Coarse-to-Fine Tokenization for Generative Recommendationhttps://blog.schoergendorfer.com/post/waymoqa-a-multi-view-visual-question-answering-dataset-for-safety-critical-reasoSecFeedde2026-09-01T11:18:06.646455+00:00WaymoQA: A Multi-View Visual Question Answering Dataset for Safety-Critical Reasoning in Autonomous Drivinghttps://blog.schoergendorfer.com/post/look-it-up-analysing-internal-web-search-capabilities-of-modern-llmsSecFeedde2026-09-01T11:18:06.601222+00:00Look It Up: Analysing Internal Web Search Capabilities of Modern LLMshttps://blog.schoergendorfer.com/post/medvision-benchmarking-quantitative-medical-image-analysisSecFeedde2026-09-01T11:03:09.305186+00:00MedVision: Benchmarking Quantitative Medical Image Analysishttps://blog.schoergendorfer.com/post/error-driven-scene-editing-for-3d-grounding-in-large-language-modelsSecFeedde2026-09-01T11:03:09.279918+00:00Error-Driven Scene Editing for 3D Grounding in Large Language Modelshttps://blog.schoergendorfer.com/post/uncertainty-makes-it-stable-curiosity-driven-quantized-mixture-of-expertsSecFeedde2026-09-01T11:03:09.258080+00:00Uncertainty Makes It Stable: Curiosity-Driven Quantized Mixture-of-Expertshttps://blog.schoergendorfer.com/post/retrofitters-pragmatists-and-activists-public-interest-litigation-for-accountablSecFeedde2026-09-01T11:03:09.227819+00:00Retrofitters, pragmatists and activists: Public interest litigation for accountable automated decision-makinghttps://blog.schoergendorfer.com/post/lemat-synth-a-multi-modal-toolbox-to-curate-broad-synthesis-procedure-databasesSecFeedde2026-09-01T11:03:09.204562+00:00LeMat-Synth: a multi-modal toolbox to curate broad synthesis procedure databases from scientific literaturehttps://blog.schoergendorfer.com/post/personalized-treatment-outcome-prediction-from-scarce-data-via-dual-channel-knowSecFeedde2026-09-01T11:03:09.182207+00:00Personalized Treatment Outcome Prediction from Scarce Data via Dual-Channel Knowledge Distillation and Adaptive Fusionhttps://blog.schoergendorfer.com/post/breaking-the-code-security-assessment-of-ai-code-agents-through-systematic-jailbSecFeedde2026-09-01T11:03:09.139700+00:00Breaking the Code: Security Assessment of AI Code Agents Through Systematic Jailbreaking Attackshttps://blog.schoergendorfer.com/post/when-llms-benchmark-themselves-deconstructing-self-bias-in-automated-evaluationSecFeedde2026-09-01T11:03:09.118584+00:00When LLMs Benchmark Themselves: Deconstructing Self-Bias in Automated Evaluationhttps://blog.schoergendorfer.com/post/operationalising-ai-regulatory-sandboxes-activities-requirements-and-technical-aSecFeedde2026-09-01T11:03:09.098339+00:00Operationalising AI Regulatory Sandboxes: Activities, Requirements, and Technical Assessment under the EU AI Acthttps://blog.schoergendorfer.com/post/talk-in-pieces-see-in-whole-disentangled-and-hierarchical-representation-learninSecFeedde2026-09-01T11:03:09.069699+00:00Talk in Pieces, See in Whole: Disentangled and Hierarchical Representation Learning in Language-based Object Detectionhttps://blog.schoergendorfer.com/post/transformer-encoder-trees-for-efficient-multilingual-machine-translation-and-speSecFeedde2026-09-01T11:03:09.044765+00:00Transformer-Encoder Trees for Efficient Multilingual Machine Translation and Speech Translationhttps://blog.schoergendorfer.com/post/magnet-coordinated-multi-agent-generation-of-synthetic-multi-turn-mental-healthSecFeedde2026-09-01T11:03:09.022035+00:00MAGneT: Coordinated Multi-Agent Generation of Synthetic Multi-Turn Mental Health Counseling Sessionshttps://blog.schoergendorfer.com/post/turning-the-spell-around-lightweight-alignment-amplification-via-rank-one-safetySecFeedde2026-09-01T11:03:08.973025+00:00Turning the Spell Around: Lightweight Alignment Amplification via Rank-One Safety Injectionhttps://blog.schoergendorfer.com/post/eegdm-learning-eeg-representation-with-latent-diffusion-modelSecFeedde2026-09-01T11:03:08.948987+00:00EEGDM: Learning EEG Representation with Latent Diffusion Modelhttps://blog.schoergendorfer.com/post/from-isolation-to-alignment-unified-lora-for-efficient-multi-task-learning-2SecFeedde2026-09-01T11:03:08.906204+00:00From Isolation to Alignment: Unified LoRA for Efficient Multi-Task Learninghttps://blog.schoergendorfer.com/post/sonicmaster-towards-controllable-all-in-one-music-restoration-and-masteringSecFeedde2026-09-01T11:03:08.884907+00:00SonicMaster: Towards Controllable All-in-One Music Restoration and Masteringhttps://blog.schoergendorfer.com/post/structsynth-dependency-graphs-as-generation-plans-for-low-data-tabular-synthesisSecFeedde2026-09-01T11:03:08.861175+00:00StructSynth: Dependency Graphs as Generation Plans for Low-Data Tabular Synthesis with Language Modelshttps://blog.schoergendorfer.com/post/proof2hybrid-automatic-mathematical-benchmark-synthesis-for-proof-centric-probleSecFeedde2026-09-01T10:47:58.139529+00:00Proof2Hybrid: Automatic Mathematical Benchmark Synthesis for Proof-Centric Problemshttps://blog.schoergendorfer.com/post/disappearing-ink-obfuscation-breaks-n-gram-code-watermarks-in-theory-and-practicSecFeedde2026-09-01T10:47:58.120178+00:00Disappearing Ink: Obfuscation Breaks N-gram Code Watermarks in Theory and Practicehttps://blog.schoergendorfer.com/post/a-foundation-model-with-multi-variate-parallel-attention-to-generate-neuronal-acSecFeedde2026-09-01T10:47:58.097454+00:00A foundation model with multi-variate parallel attention to generate neuronal activityhttps://blog.schoergendorfer.com/post/federated-learning-for-mri-based-brainage-a-multicenter-study-on-post-stroke-funSecFeedde2026-09-01T10:47:58.062997+00:00Federated Learning for MRI-based BrainAGE: a multicenter study on post-stroke functional outcome predictionhttps://blog.schoergendorfer.com/post/faa-framework-a-large-language-model-based-approach-for-credit-card-fraud-investSecFeedde2026-09-01T10:47:58.042603+00:00FAA Framework: A Large Language Model-Based Approach for Credit Card Fraud Investigationshttps://blog.schoergendorfer.com/post/evidence-grounded-trustworthy-multimodal-reasoning-and-evaluation-benchmark-in-c-3SecFeedde2026-09-01T10:47:58.026013+00:00Evidence-Grounded Trustworthy Multimodal Reasoning and Evaluation Benchmark in Complex Urban Sceneshttps://blog.schoergendorfer.com/post/understanding-automated-program-repair-agents-through-the-lens-of-traceability-aSecFeedde2026-09-01T10:47:58.004090+00:00Understanding Automated Program Repair Agents Through the Lens of Traceability: An Empirical Studyhttps://blog.schoergendorfer.com/post/mrna-design-and-optimization-with-deep-knowledge-infused-approachSecFeedde2026-09-01T10:47:57.982128+00:00mRNA Design and Optimization with Deep Knowledge-Infused Approachhttps://blog.schoergendorfer.com/post/equireg-equivariance-regularized-diffusion-for-inverse-problemsSecFeedde2026-09-01T10:47:57.956228+00:00EquiReg: Equivariance Regularized Diffusion for Inverse Problemshttps://blog.schoergendorfer.com/post/learning-composable-chains-of-thoughtSecFeedde2026-09-01T10:47:57.937761+00:00Learning Composable Chains-of-Thoughthttps://blog.schoergendorfer.com/post/watch-your-steps-dormant-adversarial-behaviors-that-activate-upon-llm-finetuningSecFeedde2026-09-01T10:47:57.918764+00:00Watch your steps: Dormant Adversarial Behaviors that Activate upon LLM Finetuninghttps://blog.schoergendorfer.com/post/regrap-llava-reasoning-enabled-graph-based-personalized-large-language-and-visioSecFeedde2026-09-01T10:47:57.898206+00:00ReGraP-LLaVA: Reasoning enabled Graph-based Personalized Large Language and Vision Assistanthttps://blog.schoergendorfer.com/post/towards-accurate-and-lightweight-peripheral-neuroblastic-tumor-diagnosis-via-conSecFeedde2026-09-01T10:47:57.878948+00:00Towards Accurate and Lightweight Peripheral Neuroblastic Tumor Diagnosis via Contrastive Multi-scale Pathological Image Analysishttps://blog.schoergendorfer.com/post/a-comprehensive-survey-on-multi-agent-cooperative-decision-making-scenarios-apprSecFeedde2026-09-01T10:47:57.857760+00:00A Comprehensive Survey on Multi-Agent Cooperative Decision-Making: Scenarios, Approaches, Challenges and Perspectiveshttps://blog.schoergendorfer.com/post/hetgb-a-comprehensive-benchmark-for-heterophilic-text-attributed-graphsSecFeedde2026-09-01T10:47:57.836837+00:00HeTGB: A Comprehensive Benchmark for Heterophilic Text-Attributed Graphshttps://blog.schoergendorfer.com/post/multimodal-large-language-models-predict-urban-safety-perception-but-encode-nonSecFeedde2026-09-01T10:47:57.817611+00:00Multimodal Large Language Models Predict Urban Safety Perception but Encode Non-Neutral Demographic Priorshttps://blog.schoergendorfer.com/post/rspo-regularized-self-play-alignment-of-large-language-modelsSecFeedde2026-09-01T10:47:57.797192+00:00RSPO: Regularized Self-Play Alignment of Large Language Modelshttps://blog.schoergendorfer.com/post/clipure-purification-in-latent-space-via-clip-for-adversarially-robust-zero-shotSecFeedde2026-09-01T10:47:57.774847+00:00CLIPure: Purification in Latent Space via CLIP for Adversarially Robust Zero-Shot Classificationhttps://blog.schoergendorfer.com/post/man-made-language-models-evaluating-llms-perpetuation-of-masculine-generics-biasSecFeedde2026-09-01T10:33:00.137657+00:00Man Made Language Models? Evaluating LLMs' Perpetuation of Masculine Generics Biashttps://blog.schoergendorfer.com/post/learning-personalized-prompts-for-healthcare-guidanceSecFeedde2026-09-01T10:33:00.114385+00:00Learning Personalized Prompts for Healthcare Guidancehttps://blog.schoergendorfer.com/post/gender-race-and-intersectional-bias-in-resume-screening-via-language-model-retriSecFeedde2026-09-01T10:33:00.075030+00:00Gender, Race, and Intersectional Bias in Resume Screening via Language Model Retrievalhttps://blog.schoergendorfer.com/post/pqmass-probabilistic-assessment-of-the-quality-of-generative-models-using-probabSecFeedde2026-09-01T10:33:00.052654+00:00PQMass: Probabilistic Assessment of the Quality of Generative Models using Probability Mass Estimationhttps://blog.schoergendorfer.com/post/sub-play-adversarial-policies-against-partially-observed-multi-agent-reinforcemeSecFeedde2026-09-01T10:33:00.026330+00:00SUB-PLAY: Adversarial Policies against Partially Observed Multi-Agent Reinforcement Learning Systemshttps://blog.schoergendorfer.com/post/general-phrase-debiaser-debiasing-masked-language-models-at-a-multi-token-levelSecFeedde2026-09-01T10:32:59.989494+00:00General Phrase Debiaser: Debiasing Masked Language Models at a Multi-Token Levelhttps://blog.schoergendorfer.com/post/logos-an-agent-harness-on-a-cross-process-busSecFeedde2026-09-01T10:32:59.942542+00:00Logos: An Agent Harness on a Cross-Process Bushttps://blog.schoergendorfer.com/post/prove2me-an-open-collaborative-platform-for-scaling-math-formalizationSecFeedde2026-09-01T10:32:59.919474+00:00Prove2Me: An Open Collaborative Platform for Scaling Math Formalizationhttps://blog.schoergendorfer.com/post/timing-aware-repurchase-prediction-for-web-scale-e-commerce-survival-models-for-2SecFeedde2026-09-01T10:32:59.893583+00:00Timing-Aware Repurchase Prediction for Web-Scale E-Commerce: Survival Models for Multi-Surface Grocery Recommendationhttps://blog.schoergendorfer.com/post/weagent-mmsearch-native-text-vision-interaction-for-multimodal-search-agentsSecFeedde2026-09-01T10:32:59.865280+00:00WeAgent-MMSearch: Native Text-Vision Interaction for Multimodal Search Agentshttps://blog.schoergendorfer.com/post/rubric-to-code-credit-assignment-for-reinforcement-learningSecFeedde2026-09-01T10:32:59.838888+00:00Rubric-to-Code Credit Assignment for Reinforcement Learninghttps://blog.schoergendorfer.com/post/realswe-a-compositional-evaluation-of-coding-agents-under-realistic-user-request-2SecFeedde2026-09-01T10:32:59.813700+00:00RealSWE: A Compositional Evaluation of Coding Agents under Realistic User Requestshttps://blog.schoergendorfer.com/post/longguard-mechanistic-analysis-and-training-free-mitigation-of-long-context-failSecFeedde2026-09-01T10:32:59.789492+00:00LongGuard: Mechanistic Analysis and Training-Free Mitigation of Long-Context Failure in Safety Guardrailshttps://blog.schoergendorfer.com/post/a-multi-modal-ai-framework-for-real-time-queue-prediction-management-and-optimis-2SecFeedde2026-09-01T10:32:59.767690+00:00A Multi-Modal AI Framework for Real-Time Queue Prediction, Management and Optimisation in Intelligent Border Control Systemshttps://blog.schoergendorfer.com/post/progrouter-online-progress-guided-orchestration-for-multi-agent-llm-workflows-unSecFeedde2026-09-01T10:32:59.726019+00:00ProgRouter: Online Progress-Guided Orchestration for Multi-Agent LLM Workflows under Quality-Cost Tradeoffshttps://blog.schoergendorfer.com/post/candidate-supply-and-answer-selection-shape-the-value-of-llm-judging-in-multi-agSecFeedde2026-09-01T10:32:59.704825+00:00Candidate supply and answer selection shape the value of LLM judging in multi-agent systemshttps://blog.schoergendorfer.com/post/repair-or-resample-rethinking-failure-debugging-in-llm-multi-agent-systemsSecFeedde2026-09-01T10:32:59.682949+00:00Repair or Resample? Rethinking Failure Debugging in LLM Multi-Agent Systemshttps://blog.schoergendorfer.com/post/towards-reliable-generalizable-and-specific-in-context-knowledge-editing-via-mul-2SecFeedde2026-09-01T10:17:59.828831+00:00Towards Reliable, Generalizable, and Specific In-Context Knowledge Editing via Multi-Objective Reinforcement Learninghttps://blog.schoergendorfer.com/post/post-graph-rag-a-postgresql-native-bi-temporal-graph-rag-engine-with-temporal-grSecFeedde2026-09-01T10:17:59.804588+00:00post-graph-rag: A PostgreSQL-Native Bi-Temporal Graph RAG Engine with Temporal Grounding at Synthesishttps://blog.schoergendorfer.com/post/simguide-typed-multi-context-user-representations-for-preference-conditioned-ageSecFeedde2026-09-01T10:17:59.776255+00:00SimGuide: Typed Multi-Context User Representations for Preference-Conditioned Agent Planninghttps://blog.schoergendorfer.com/post/when-seeing-is-not-enough-benchmarking-interactive-visual-grounding-in-lvlms-2SecFeedde2026-09-01T10:17:59.756600+00:00When Seeing Is Not Enough: Benchmarking Interactive Visual Grounding in LVLMshttps://blog.schoergendorfer.com/post/robustness-analysis-of-agentic-ai-to-inconsistent-and-incomplete-tool-responses-2SecFeedde2026-09-01T10:17:59.712829+00:00Robustness Analysis of Agentic AI to Inconsistent and Incomplete Tool Responseshttps://blog.schoergendorfer.com/post/execrubrics-executable-tool-augmented-rubrics-for-verifiable-and-efficient-long-3SecFeedde2026-09-01T10:17:59.688797+00:00ExecRubrics: Executable Tool-Augmented Rubrics for Verifiable and Efficient Long-Form Evaluationhttps://blog.schoergendorfer.com/post/audita-certified-auditing-and-causal-attribution-of-adverse-outcomes-in-autonomo-2SecFeedde2026-09-01T10:17:59.659202+00:00AUDITA: certified auditing and causal attribution of adverse outcomes in autonomous multi-agent systemshttps://blog.schoergendorfer.com/post/is-multimodal-speculative-decoding-ready-for-diffusion-based-parallel-drafting-a-2SecFeedde2026-09-01T10:17:59.633589+00:00Is Multimodal Speculative Decoding Ready for Diffusion-Based Parallel Drafting? A Survey and Empirical Diagnosishttps://blog.schoergendorfer.com/post/sapo-single-rollout-autoregressive-policy-optimization-for-agentic-reinforcement-2SecFeedde2026-09-01T10:17:59.603730+00:00SAPO: Single-Rollout Autoregressive Policy Optimization for Agentic Reinforcement Learninghttps://blog.schoergendorfer.com/post/rtpo-reverse-turn-policy-optimization-for-stabilizing-agentic-rl-training-2SecFeedde2026-09-01T10:17:59.578100+00:00RTPO: Reverse-Turn Policy Optimization for Stabilizing Agentic RL Traininghttps://blog.schoergendorfer.com/post/beyond-the-trace-coupling-an-interpretable-reasoning-state-readout-to-native-moe-2SecFeedde2026-09-01T10:17:59.554376+00:00Beyond the Trace: Coupling an Interpretable Reasoning-State Readout to Native MoE Routinghttps://blog.schoergendorfer.com/post/constitutive-priors-for-machine-intelligence-a-legitimacy-theory-of-the-artifici-2SecFeedde2026-09-01T10:17:59.529279+00:00Constitutive Priors for Machine Intelligence: A Legitimacy Theory of the Artificial Physical Worldhttps://blog.schoergendorfer.com/post/evo-harness-context-to-harness-skill-compilation-for-self-evolving-agents-2SecFeedde2026-09-01T10:17:59.503925+00:00Evo-Harness: Context-to-Harness Skill Compilation for Self-Evolving Agentshttps://blog.schoergendorfer.com/post/renormalising-generative-models-for-active-inference-foundations-derivations-and-2SecFeedde2026-09-01T10:17:59.460737+00:00Renormalising Generative Models for Active Inference: Foundations, Derivations, and Verificationhttps://blog.schoergendorfer.com/post/psychoagent-an-affect-sensitive-cognitive-architecture-for-conflict-aware-memory-2SecFeedde2026-09-01T10:17:59.433878+00:00PsychoAgent: An Affect-Sensitive Cognitive Architecture for Conflict-Aware Memory in LLM Agentshttps://blog.schoergendorfer.com/post/the-optimizer-is-the-agent-reasoning-driven-search-across-prompts-programs-and-m-2SecFeedde2026-09-01T10:17:59.404452+00:00The Optimizer Is the Agent: Reasoning-Driven Search across Prompts, Programs, and ML Workflowshttps://blog.schoergendorfer.com/post/nxn-e-valuation-hypothesis-certification-via-a-conformal-crt-null-2SecFeedde2026-09-01T10:17:59.380364+00:00NxN E-valuation: Hypothesis Certification via a Conformal CRT Nullhttps://blog.schoergendorfer.com/post/unequal-verdicts-investigating-gender-bias-in-llm-based-fake-news-detection-2SecFeedde2026-09-01T10:17:59.349744+00:00Unequal Verdicts: Investigating Gender Bias in LLM-Based Fake News Detectionhttps://blog.schoergendorfer.com/post/messier-a-high-resolution-corpus-for-cross-benchmark-agent-evaluation-2SecFeedde2026-09-01T10:03:09.712364+00:00Messier: A High-Resolution Corpus for Cross-Benchmark Agent Evaluationhttps://blog.schoergendorfer.com/post/instruction-tuned-language-models-cannot-sample-from-distributions-they-can-desc-2SecFeedde2026-09-01T10:03:09.692213+00:00Instruction-Tuned Language Models Cannot Sample from Distributions They Can Describehttps://blog.schoergendorfer.com/post/masked-distillation-internalizing-the-chain-of-thought-in-language-models-2SecFeedde2026-09-01T10:03:09.669129+00:00Masked Distillation: Internalizing the Chain-of-Thought in Language Modelshttps://blog.schoergendorfer.com/post/dynamicmcpbench-a-trace-grounded-effect-scored-benchmark-for-llm-agents-over-liv-2SecFeedde2026-09-01T10:03:09.641608+00:00DynamicMCPBench: A Trace-Grounded, Effect-Scored Benchmark for LLM Agents over Live MCP Servershttps://blog.schoergendorfer.com/post/evosql-memory-augmented-critic-generator-co-evolution-for-text-to-sql-2SecFeedde2026-09-01T10:03:09.616820+00:00EvoSQL: Memory-Augmented Critic-Generator Co-Evolution for Text-to-SQLhttps://blog.schoergendorfer.com/post/worldcuparena-fine-grained-evaluation-of-language-models-and-deep-research-agent-2SecFeedde2026-09-01T10:03:09.592982+00:00WorldCupArena: Fine-Grained Evaluation of Language Models and Deep-Research Agents on Football Forecastinghttps://blog.schoergendorfer.com/post/can-agentic-trading-systems-pay-for-their-own-intelligence-2SecFeedde2026-09-01T10:03:09.568631+00:00Can Agentic Trading Systems Pay for Their Own Intelligence?https://blog.schoergendorfer.com/post/where-knowledge-and-authority-sit-changes-what-an-agent-benchmark-can-resolveSecFeedde2026-09-01T10:03:09.528180+00:00Where Knowledge and Authority Sit Changes What an Agent Benchmark Can Resolvehttps://blog.schoergendorfer.com/post/mnemosyne-agentic-transaction-processing-for-validating-and-repairing-ai-generat-2SecFeedde2026-09-01T10:03:09.505555+00:00Mnemosyne: Agentic Transaction Processing for Validating and Repairing AI-generated Workflowshttps://blog.schoergendorfer.com/post/compass-grounding-composition-intent-guidance-in-unified-multimodal-modelsSecFeedde2026-09-01T10:03:09.479715+00:00COMPASS: Grounding Composition-Intent Guidance in Unified Multimodal Modelshttps://blog.schoergendorfer.com/post/rift-bench-dynamic-red-teaming-for-agentic-ai-systemsSecFeedde2026-09-01T10:03:09.455823+00:00RIFT-Bench: Dynamic Red-teaming For Agentic AI Systemshttps://blog.schoergendorfer.com/post/dart-draft-agreement-routing-for-training-free-adaptive-thinking-budgets-in-hybrSecFeedde2026-09-01T10:03:09.433575+00:00DART: Draft-Agreement Routing for Training-Free Adaptive Thinking Budgets in Hybrid Reasoning Modelshttps://blog.schoergendorfer.com/post/in-llm-reasoning-there-is-irrationality-on-top-of-value-misalignmentSecFeedde2026-09-01T10:03:09.410253+00:00In LLM Reasoning, there is Irrationality on top of Value Misalignmenthttps://blog.schoergendorfer.com/post/scaffoldagent-utility-guided-dynamic-outline-optimization-for-open-ended-deep-reSecFeedde2026-09-01T10:03:09.372834+00:00ScaffoldAgent: Utility-Guided Dynamic Outline Optimization for Open-Ended Deep Researchhttps://blog.schoergendorfer.com/post/a-homotopy-type-theoretic-generalization-of-neurosymbolic-inferenceSecFeedde2026-09-01T10:03:09.333081+00:00A homotopy-type-theoretic generalization of neurosymbolic inferencehttps://blog.schoergendorfer.com/post/beyond-helpfulness-a-teaching-over-solving-diagnostic-for-measuring-educationalSecFeedde2026-09-01T09:47:35.753660+00:00Beyond Helpfulness: A Teaching-over-Solving Diagnostic for Measuring Educational Impact in LLM Tutorshttps://blog.schoergendorfer.com/post/itercad-an-iterative-multimodal-agent-for-visually-grounded-cad-generation-and-eSecFeedde2026-09-01T09:47:35.726572+00:00IterCAD: An Iterative Multimodal Agent for Visually-Grounded CAD Generation and Editinghttps://blog.schoergendorfer.com/post/from-agi-to-asiSecFeedde2026-09-01T09:47:35.703935+00:00From AGI to ASIhttps://blog.schoergendorfer.com/post/foresightsafety-sagea-fully-automated-scenario-generation-and-safety-evaluation-2SecFeedde2026-09-01T09:47:35.678892+00:00ForesightSafety-SAGE:A Fully Automated Scenario Generation and Safety Evaluation Framework for LLM Agentshttps://blog.schoergendorfer.com/post/how-small-can-you-go-lora-fine-tuning-270m-8b-models-for-merchant-information-ex-2SecFeedde2026-09-01T09:47:35.654045+00:00How Small Can You Go? LoRA Fine-Tuning 270M-8B Models for Merchant Information Extraction in Financial Transactionshttps://blog.schoergendorfer.com/post/crowdmath-a-dataset-of-crowdsourced-mathematical-research-discussionsSecFeedde2026-09-01T09:47:35.629518+00:00CrowdMath: A Dataset of Crowdsourced Mathematical Research Discussionshttps://blog.schoergendorfer.com/post/evolving-agents-in-the-dark-retrospective-harness-optimization-via-self-preferenSecFeedde2026-09-01T09:47:35.610156+00:00Evolving Agents in the Dark: Retrospective Harness Optimization via Self-Preferencehttps://blog.schoergendorfer.com/post/handoff-debt-the-rediscovery-cost-when-coding-agents-take-over-interrupted-tasksSecFeedde2026-09-01T09:47:35.574106+00:00Handoff Debt: The Rediscovery Cost When Coding Agents Take Over Interrupted Taskshttps://blog.schoergendorfer.com/post/self-correction-can-amplify-hallucinations-fact-level-repair-with-graph-based-evSecFeedde2026-09-01T09:47:35.535812+00:00Self-Correction Can Amplify Hallucinations: Fact-Level Repair with Graph-Based Evidence Routing in Multimodal Generationhttps://blog.schoergendorfer.com/post/citation-closure-retrieval-and-per-rule-attribution-for-real-world-regulatory-coSecFeedde2026-09-01T09:47:35.477345+00:00Citation-Closure Retrieval and Per-Rule Attribution for Real-World Regulatory Compliance Question Answeringhttps://blog.schoergendorfer.com/post/adopt-neq-adapt-longitudinal-analyses-of-llm-conversations-in-the-wildSecFeedde2026-09-01T09:47:35.453030+00:00Adopt $\neq$ Adapt: Longitudinal Analyses of LLM Conversations in the Wildhttps://blog.schoergendorfer.com/post/zipping-the-thought-when-and-how-compressed-reasoning-data-works-in-llm-post-traSecFeedde2026-09-01T09:47:35.429440+00:00Zipping the Thought: When and How Compressed Reasoning Data Works in LLM Post-Traininghttps://blog.schoergendorfer.com/post/asynctool-evaluating-the-asynchronous-function-calling-capability-under-multi-taSecFeedde2026-09-01T09:47:35.408279+00:00AsyncTool: Evaluating the Asynchronous Function Calling Capability under Multi-Task Scenarioshttps://blog.schoergendorfer.com/post/portbench-a-correlation-aware-full-pipeline-benchmark-for-llm-driven-portfolio-m-2SecFeedde2026-09-01T09:47:35.390131+00:00PortBench: A Correlation-Aware, Full-Pipeline Benchmark for LLM-Driven Portfolio Managementhttps://blog.schoergendorfer.com/post/veritrace-evolving-mental-models-for-deep-research-agentsSecFeedde2026-09-01T09:47:35.370451+00:00VeriTrace: Evolving Mental Models for Deep Research Agentshttps://blog.schoergendorfer.com/post/agent-as-peer-debriefer-a-multi-agent-framework-with-perspective-based-refinemenSecFeedde2026-09-01T09:47:35.344144+00:00Agent-as-Peer-Debriefer: A Multi-Agent Framework with Perspective-Based Refinement for Qualitative Analysishttps://blog.schoergendorfer.com/post/summoning-the-oracle-to-slay-it-mitigating-look-ahead-bias-in-financial-backtestSecFeedde2026-09-01T09:32:43.022363+00:00Summoning the Oracle to Slay It: Mitigating Look-Ahead Bias in Financial Backtesting with Large Language Modelshttps://blog.schoergendorfer.com/post/sciatlas-a-computable-atlas-of-science-for-knowledge-grounded-ai-researchSecFeedde2026-09-01T09:32:42.999802+00:00SciAtlas: A Computable Atlas of Science for Knowledge-Grounded AI Researchhttps://blog.schoergendorfer.com/post/useful-memories-become-faulty-when-continuously-updated-by-llmsSecFeedde2026-09-01T09:32:42.980257+00:00Useful Memories Become Faulty When Continuously Updated by LLMshttps://blog.schoergendorfer.com/post/skillret-a-large-scale-benchmark-for-skill-retrieval-in-llm-agentsSecFeedde2026-09-01T09:32:42.956530+00:00SkillRet: A Large-Scale Benchmark for Skill Retrieval in LLM Agentshttps://blog.schoergendorfer.com/post/first-order-efficiency-for-probabilistic-value-estimation-via-a-statistical-viewSecFeedde2026-09-01T09:32:42.927082+00:00First-Order Efficiency for Probabilistic Value Estimation via A Statistical Viewpointhttps://blog.schoergendorfer.com/post/training-and-agentic-inference-strategies-for-llm-based-manim-animation-generatiSecFeedde2026-09-01T09:32:42.906461+00:00Training and Agentic Inference Strategies for LLM-based Manim Animation Generationhttps://blog.schoergendorfer.com/post/reactbench-a-benchmark-for-topological-reasoning-in-mllms-on-chemical-reaction-d-2SecFeedde2026-09-01T09:32:42.884281+00:00ReactBench: A Benchmark for Topological Reasoning in MLLMs on Chemical Reaction Diagramshttps://blog.schoergendorfer.com/post/webxskill-skill-learning-for-autonomous-web-agentsSecFeedde2026-09-01T09:32:42.861055+00:00WebXSkill: Skill Learning for Autonomous Web Agentshttps://blog.schoergendorfer.com/post/a-multi-agent-human-llm-collaborative-framework-for-closed-loop-scientific-literSecFeedde2026-09-01T09:32:42.822361+00:00A Multi-Agent Human-LLM Collaborative Framework for Closed-Loop Scientific Literature Summarizationhttps://blog.schoergendorfer.com/post/view-oriented-conversation-compiler-for-agent-trace-analysisSecFeedde2026-09-01T09:32:42.799943+00:00View-oriented Conversation Compiler for Agent Trace Analysishttps://blog.schoergendorfer.com/post/peoplesearchbench-evaluating-ai-powered-people-search-platforms-with-criteria-grSecFeedde2026-09-01T09:32:42.777777+00:00PeopleSearchBench: Evaluating AI-Powered People Search Platforms with Criteria-Grounded Verificationhttps://blog.schoergendorfer.com/post/intrygue-induction-aware-entropy-gating-for-reliable-rag-uncertainty-estimation-2SecFeedde2026-09-01T09:32:42.741125+00:00INTRYGUE: Induction-Aware Entropy Gating for Reliable RAG Uncertainty Estimationhttps://blog.schoergendorfer.com/post/retrieval-augmented-llm-agents-learning-to-learn-from-experienceSecFeedde2026-09-01T09:32:42.717274+00:00Retrieval-Augmented LLM Agents: Learning to Learn from Experiencehttps://blog.schoergendorfer.com/post/unified-mas-universally-generating-domain-specific-nodes-for-empowering-automatiSecFeedde2026-09-01T09:32:42.695906+00:00Unified-MAS: Universally Generating Domain-Specific Nodes for Empowering Automatic Multi-Agent Systemshttps://blog.schoergendorfer.com/post/agentrx-diagnosing-ai-agent-failures-from-execution-trajectoriesSecFeedde2026-09-01T09:32:42.657998+00:00AgentRx: Diagnosing AI Agent Failures from Execution Trajectorieshttps://blog.schoergendorfer.com/post/beyond-dense-states-sparse-transcoders-as-causally-testable-operators-for-llm-laSecFeedde2026-09-01T09:32:42.637103+00:00Beyond Dense States: Sparse Transcoders as Causally Testable Operators for LLM Latent Reasoninghttps://blog.schoergendorfer.com/post/reasoning-and-tool-use-compete-in-agentic-rlfrom-quantifying-interference-to-disSecFeedde2026-09-01T09:32:42.616388+00:00Reasoning and Tool-use Compete in Agentic RL:From Quantifying Interference to Disentangled Tuninghttps://blog.schoergendorfer.com/post/humanstudy-bench-towards-ai-agent-design-for-participant-simulationSecFeedde2026-09-01T09:17:25.134491+00:00HumanStudy-Bench: Towards AI Agent Design for Participant Simulationhttps://blog.schoergendorfer.com/post/defining-operational-conditions-for-safety-critical-ai-based-systems-from-dataSecFeedde2026-09-01T09:17:25.114069+00:00Defining Operational Conditions for Safety-Critical AI-Based Systems from Datahttps://blog.schoergendorfer.com/post/shardmemo-scope-before-routing-for-agentic-memory-retrievalSecFeedde2026-09-01T09:17:25.093821+00:00ShardMemo: Scope-Before-Routing for Agentic Memory Retrievalhttps://blog.schoergendorfer.com/post/real-time-deadlines-reveal-fragile-temporal-adaptation-in-llm-strategic-dialogueSecFeedde2026-09-01T09:17:25.054849+00:00Real-Time Deadlines Reveal Fragile Temporal Adaptation in LLM Strategic Dialogueshttps://blog.schoergendorfer.com/post/spring-continual-llm-personalization-via-selective-parametric-adaptation-and-ret-2SecFeedde2026-09-01T09:17:25.032886+00:00SPRInG: Continual LLM Personalization via Selective Parametric Adaptation and Retrieval-Interpolated Generationhttps://blog.schoergendorfer.com/post/controllable-memory-usage-balancing-anchoring-and-innovation-in-long-term-humanSecFeedde2026-09-01T09:17:25.010851+00:00Controllable Memory Usage: Balancing Anchoring and Innovation in Long-Term Human-Agent Interactionhttps://blog.schoergendorfer.com/post/mobiledreamer-generative-sketch-world-model-for-gui-agentSecFeedde2026-09-01T09:17:24.991617+00:00MobileDreamer: Generative Sketch World Model for GUI Agenthttps://blog.schoergendorfer.com/post/gui-pra-process-reward-agent-for-gui-tasksSecFeedde2026-09-01T09:17:24.954565+00:00GUI-PRA: Process Reward Agent for GUI Taskshttps://blog.schoergendorfer.com/post/reasoning-or-rambling-exploring-the-effect-of-thinking-on-agent-persuasion-2SecFeedde2026-09-01T09:17:24.931256+00:00Reasoning or Rambling? Exploring the Effect of Thinking on Agent Persuasionhttps://blog.schoergendorfer.com/post/spade-a-large-language-model-framework-for-soil-moisture-pattern-recognition-andSecFeedde2026-09-01T09:17:24.895561+00:00SPADE: A Large Language Model Framework for Soil Moisture Pattern Recognition and Anomaly Detection in Precision Agriculturehttps://blog.schoergendorfer.com/post/beyond-pixels-exploring-dom-downsampling-for-llm-based-web-agents-2SecFeedde2026-09-01T09:17:24.877272+00:00Beyond Pixels: Exploring DOM Downsampling for LLM-Based Web Agentshttps://blog.schoergendorfer.com/post/assetopsbench-benchmarking-ai-agents-for-task-automation-in-industrial-asset-opeSecFeedde2026-09-01T09:17:24.857345+00:00AssetOpsBench: Benchmarking AI Agents for Task Automation in Industrial Asset Operations and Maintenancehttps://blog.schoergendorfer.com/post/a-mental-model-based-framework-of-trustSecFeedde2026-09-01T09:17:24.834715+00:00A Mental Model Based Framework of Trusthttps://blog.schoergendorfer.com/post/sun-persistent-programs-for-language-grounded-control-to-learning-to-real-policiSecFeedde2026-09-01T09:17:24.812593+00:00SUN: Persistent Programs For Language-Grounded Control-to-Learning-to-Real Policieshttps://blog.schoergendorfer.com/post/auditing-anonymous-ai-models-a-four-stage-protocol-for-black-box-identity-verifiSecFeedde2026-09-01T09:17:24.790630+00:00Auditing Anonymous AI Models: A Four-Stage Protocol for Black-Box Identity Verificationhttps://blog.schoergendorfer.com/post/llm-post-training-as-brownfield-maintenance-an-industrial-perspective-on-datawarSecFeedde2026-09-01T09:17:24.771821+00:00LLM Post-Training as Brownfield Maintenance: An Industrial Perspective on Dataware Engineeringhttps://blog.schoergendorfer.com/post/learning-to-evaluate-before-improving-automatic-rubric-induction-for-automatic-rSecFeedde2026-09-01T09:17:24.750594+00:00Learning to Evaluate Before Improving: Automatic Rubric Induction for Automatic Research Agentshttps://blog.schoergendorfer.com/post/one-note-in-three-a-verified-census-of-three-deployed-ai-scribes-and-the-instrumSecFeedde2026-09-01T09:02:39.208619+00:00One note in three: a verified census of three deployed AI scribes, and the instrument that counted ithttps://blog.schoergendorfer.com/post/llm-judges-verify-presence-not-absence-omission-blindness-in-ai-clinical-notes-aSecFeedde2026-09-01T09:02:39.186021+00:00LLM Judges Verify Presence, Not Absence: Omission Blindness in AI Clinical Notes and What Recovers Ithttps://blog.schoergendorfer.com/post/stick-to-what-you-know-a-study-of-knowledge-aligned-supervised-fine-tuningSecFeedde2026-09-01T09:02:39.162295+00:00Stick to What You Know: A Study of Knowledge-Aligned Supervised Fine-Tuninghttps://blog.schoergendorfer.com/post/mr-jepa-a-general-purpose-video-foundation-model-for-cardiac-mriSecFeedde2026-09-01T09:02:39.120906+00:00MR-JEPA: A General Purpose Video Foundation Model for Cardiac MRIhttps://blog.schoergendorfer.com/post/cojepa-combining-contrastive-learning-and-jepa-for-global-local-music-representaSecFeedde2026-09-01T09:02:39.099271+00:00CoJEPA: Combining Contrastive Learning and JEPA for Global-Local Music Representationshttps://blog.schoergendorfer.com/post/cogevol-towards-efficient-and-reliable-learning-environment-generationSecFeedde2026-09-01T09:02:39.073510+00:00CogEvol: Towards Efficient and Reliable Learning Environment Generationhttps://blog.schoergendorfer.com/post/a-universal-context-reuse-layer-for-cross-model-kv-sharingSecFeedde2026-09-01T09:02:39.049535+00:00A Universal Context-Reuse Layer for Cross-Model KV Sharinghttps://blog.schoergendorfer.com/post/loci-a-locator-critic-with-refinement-loopSecFeedde2026-09-01T09:02:39.027395+00:00LOCI: A Locator-Critic with Refinement Loophttps://blog.schoergendorfer.com/post/taking-the-whys-seriously-limitations-of-counterfactual-explanations-in-justificSecFeedde2026-09-01T09:02:39.007029+00:00Taking the Whys Seriously: Limitations of Counterfactual Explanations in Justification and Recoursehttps://blog.schoergendorfer.com/post/musgu-toward-a-musician-centered-evaluation-framework-and-discovery-tool-for-genSecFeedde2026-09-01T09:02:38.985275+00:00MusGU+: Toward a Musician-Centered Evaluation Framework and Discovery Tool for Generative Music AIhttps://blog.schoergendorfer.com/post/lightnav-0-eliciting-vlm-spatial-intelligence-for-generalist-embodied-navigationSecFeedde2026-09-01T09:02:38.964304+00:00LightNav-0: Eliciting VLM Spatial Intelligence for Generalist Embodied Navigationhttps://blog.schoergendorfer.com/post/stride-k-subsampling-train-free-audio-token-reduction-for-whisperSecFeedde2026-09-01T09:02:38.942013+00:00Stride-k Subsampling: Train-Free Audio Token Reduction for Whisperhttps://blog.schoergendorfer.com/post/safety-screening-for-voltage-control-in-active-distribution-grids-via-distributiSecFeedde2026-09-01T09:02:38.915028+00:00Safety Screening for Voltage Control in Active Distribution Grids via Distributionally Robust Conformal Screeninghttps://blog.schoergendorfer.com/post/towards-stream-learning-on-embedded-systems-benchmarking-the-memory-consumptionSecFeedde2026-09-01T09:02:38.891367+00:00Towards Stream Learning on Embedded Systems: Benchmarking the Memory Consumption of Stream Learning Methodshttps://blog.schoergendorfer.com/post/personas-differ-from-native-language-generation-language-pathways-shape-llm-inteSecFeedde2026-09-01T09:02:38.844839+00:00Personas Differ from Native-Language Generation: Language Pathways Shape LLM Interpersonal Advicehttps://blog.schoergendorfer.com/post/exponential-random-graph-models-with-soft-clique-constraintsSecFeedde2026-09-01T09:02:38.820473+00:00Exponential random graph models with soft clique constraintshttps://blog.schoergendorfer.com/post/tami-temporally-aligned-missingness-aware-and-interpretable-multimodal-fusion-foSecFeedde2026-09-01T09:02:38.787572+00:00TAMI: Temporally Aligned, Missingness-Aware, and Interpretable Multimodal Fusion for Mental Health Assessment in Older Adults with Mild Cognitive Impairmenthttps://blog.schoergendorfer.com/post/pretrained-curriculum-tuned-and-ensembled-a-tracer-aware-interactive-segmentatioSecFeedde2026-09-01T08:47:35.800705+00:00Pretrained, Curriculum-Tuned, and Ensembled: A Tracer-Aware Interactive Segmentation Pipeline for AutoPET Vhttps://blog.schoergendorfer.com/post/reliable-benchmarking-of-artifact-detection-in-computational-pathology-a-reproduSecFeedde2026-09-01T08:47:35.779715+00:00Reliable Benchmarking of Artifact Detection in Computational Pathology: A Reproducibility and Uncertainty Analysishttps://blog.schoergendorfer.com/post/lucida-parse-generate-and-place-for-composable-real-to-sim-scene-modelingSecFeedde2026-09-01T08:47:35.751730+00:00Lucida: Parse, Generate, and Place for Composable Real-to-Sim Scene Modelinghttps://blog.schoergendorfer.com/post/a-composition-aware-pretraining-framework-for-geospatial-foundation-modelsSecFeedde2026-09-01T08:47:35.723363+00:00A Composition-Aware Pretraining Framework for Geospatial Foundation Modelshttps://blog.schoergendorfer.com/post/aggregate-disambiguation-systemsSecFeedde2026-09-01T08:47:35.701258+00:00Aggregate Disambiguation Systemshttps://blog.schoergendorfer.com/post/conjoint-audio-to-spikes-encoding-and-processing-for-efficient-neuromorphic-speeSecFeedde2026-09-01T08:47:35.678997+00:00Conjoint Audio-to-Spikes Encoding and Processing for Efficient Neuromorphic Speech Recognitionhttps://blog.schoergendorfer.com/post/on-the-prospects-of-dynamic-llm-conversations-in-software-developmentSecFeedde2026-09-01T08:47:35.654576+00:00On the Prospects of Dynamic LLM Conversations in Software Developmenthttps://blog.schoergendorfer.com/post/calibrating-small-language-models-for-claim-check-worthiness-detectionSecFeedde2026-09-01T08:47:35.631687+00:00Calibrating Small Language Models for Claim Check-Worthiness Detectionhttps://blog.schoergendorfer.com/post/railgen-improving-railway-intrusion-detection-via-agent-guided-small-scale-foreiSecFeedde2026-09-01T08:47:35.588633+00:00RailGen: Improving Railway Intrusion Detection via Agent-Guided Small-Scale Foreign Object Generationhttps://blog.schoergendorfer.com/post/baitbench-measuring-agent-reward-hacking-with-optional-shortcuts-planted-in-ml-tSecFeedde2026-09-01T08:47:35.562428+00:00BAITBENCH: Measuring Agent Reward Hacking with Optional Shortcuts Planted in ML Taskshttps://blog.schoergendorfer.com/post/railsyn-diagnosis-guided-image-generation-for-traceable-data-completion-in-railwSecFeedde2026-09-01T08:47:35.542282+00:00RailSyn: Diagnosis-Guided Image Generation for Traceable Data Completion in Railway Foreign Object Detectionhttps://blog.schoergendorfer.com/post/singprobe-technical-reportSecFeedde2026-09-01T08:47:35.520797+00:00SingProbe Technical Reporthttps://blog.schoergendorfer.com/post/an-agentic-retrobiosynthesis-framework-with-learned-frontier-selectionSecFeedde2026-09-01T08:47:35.494434+00:00An Agentic Retrobiosynthesis Framework with Learned Frontier Selectionhttps://blog.schoergendorfer.com/post/learning-dynamics-of-logits-debiasing-for-long-tailed-semi-supervised-learningSecFeedde2026-09-01T08:47:35.474592+00:00Learning Dynamics of Logits Debiasing for Long-Tailed Semi-Supervised Learninghttps://blog.schoergendorfer.com/post/learning-materials-properties-from-scarce-labels-and-unlabeled-crystalsSecFeedde2026-09-01T08:47:35.454623+00:00Learning Materials Properties from Scarce Labels and Unlabeled Crystalshttps://blog.schoergendorfer.com/post/compass-collaborative-molecular-property-prediction-via-adaptive-small-large-modSecFeedde2026-09-01T08:47:35.414113+00:00CoMPASS: Collaborative Molecular Property Prediction via Adaptive Small-Large Model Synergyhttps://blog.schoergendorfer.com/post/fine-grained-multi-image-object-hallucination-benchmarkSecFeedde2026-09-01T08:47:35.392251+00:00Fine-Grained Multi Image Object Hallucination Benchmarkhttps://blog.schoergendorfer.com/post/big-sure-bipartite-graph-for-semantic-uncertainty-and-reliability-estimation-ofSecFeedde2026-09-01T08:47:35.365823+00:00BiG-SURE - Bipartite Graph for Semantic Uncertainty and Reliability Estimation of LLMshttps://blog.schoergendorfer.com/post/gmts-gradient-magnitude-based-token-selection-improves-rlvr-training-for-llm-reaSecFeedde2026-09-01T08:32:58.891344+00:00GMTS: Gradient Magnitude-based Token Selection Improves RLVR Training for LLM Reasoninghttps://blog.schoergendorfer.com/post/cost-efficient-active-learning-for-referring-image-segmentation-and-groundingSecFeedde2026-09-01T08:32:58.870007+00:00Cost-efficient Active Learning for Referring Image Segmentation and Groundinghttps://blog.schoergendorfer.com/post/hidden-threat-in-synthetic-data-covert-targeted-bias-injection-through-benign-teSecFeedde2026-09-01T08:32:58.845712+00:00Hidden Threat in Synthetic Data: Covert Targeted Bias Injection through Benign Texthttps://blog.schoergendorfer.com/post/reading-the-news-adapting-large-language-models-to-swedish-journalism-through-coSecFeedde2026-09-01T08:32:58.820738+00:00Reading the News: Adapting Large Language Models to Swedish Journalism Through Continued Pre-Traininghttps://blog.schoergendorfer.com/post/generative-retrieval-for-e-commerce-jointly-learning-embedding-and-codebook-withSecFeedde2026-09-01T08:32:58.797784+00:00Generative Retrieval for E-commerce: Jointly Learning Embedding and Codebook with Same Product Clusterhttps://blog.schoergendorfer.com/post/diffsac-diffusion-guided-sampling-for-consensus-based-robust-estimationSecFeedde2026-09-01T08:32:58.778357+00:00DiffSAC: Diffusion-guided Sampling for Consensus-based Robust Estimationhttps://blog.schoergendorfer.com/post/collapsibility-of-performance-metrics-in-clinical-predictive-aiSecFeedde2026-09-01T08:32:58.740994+00:00Collapsibility of Performance Metrics in Clinical Predictive AIhttps://blog.schoergendorfer.com/post/q-strata-hierarchical-bit-allocation-for-mixed-precision-quantization-of-mixtureSecFeedde2026-09-01T08:32:58.701364+00:00Q-Strata: Hierarchical Bit Allocation for Mixed-Precision Quantization of Mixture-of-Experts LLMshttps://blog.schoergendorfer.com/post/preference-shapes-relevance-cross-component-hierarchical-semantic-alignment-forSecFeedde2026-09-01T08:32:58.681503+00:00Preference Shapes Relevance: Cross-component Hierarchical Semantic Alignment for Personalized Generative Retrievalhttps://blog.schoergendorfer.com/post/trajectory-initialized-neural-double-q-routing-for-large-scale-overhead-hoist-trSecFeedde2026-09-01T08:32:58.637909+00:00Trajectory-Initialized Neural Double Q-Routing for Large-Scale Overhead Hoist Transport Systemshttps://blog.schoergendorfer.com/post/lot-machine-multimodal-lot-extraction-from-auction-catalogsSecFeedde2026-09-01T08:32:58.615194+00:00Lot Machine: Multimodal Lot Extraction from Auction Catalogshttps://blog.schoergendorfer.com/post/tensor-methods-for-language-models-from-token-representation-to-training-adaptatSecFeedde2026-09-01T08:32:58.592576+00:00Tensor Methods for Language Models: From Token Representation to Training, Adaptation, Inference, Compression, and Interpretabilityhttps://blog.schoergendorfer.com/post/measuring-memory-and-generalization-as-separable-geometric-channels-the-topo2-frSecFeedde2026-09-01T08:32:58.568553+00:00Measuring Memory and Generalization as Separable Geometric Channels: The Topo^2 Frameworkhttps://blog.schoergendorfer.com/post/enhancing-low-resource-language-reasoning-via-high-resource-language-feature-traSecFeedde2026-09-01T08:32:58.523326+00:00Enhancing Low-Resource Language Reasoning via High-Resource Language Feature Transferhttps://blog.schoergendorfer.com/post/lies-we-can-see-joint-verbal-and-non-verbal-deception-by-vlm-agents-in-embodiedSecFeedde2026-09-01T08:32:58.499780+00:00Lies We Can See: Joint Verbal and Non-Verbal Deception by VLM Agents in Embodied Social Interactionshttps://blog.schoergendorfer.com/post/towards-cognitive-process-aware-proactive-writing-supportSecFeedde2026-09-01T08:32:58.475774+00:00Towards Cognitive Process-Aware Proactive Writing Supporthttps://blog.schoergendorfer.com/post/whole-slide-image-analysis-under-realistic-few-shot-annotation-protocolsSecFeedde2026-09-01T08:32:58.431897+00:00Whole-Slide Image Analysis under Realistic Few-Shot Annotation Protocolshttps://blog.schoergendorfer.com/post/imagecas-x-a-dataset-and-benchmark-for-coronary-artery-segmentation-and-centerliSecFeedde2026-09-01T08:17:50.127481+00:00ImageCAS-X: a dataset and benchmark for coronary artery segmentation and centerline extraction in coronary CT angiographyhttps://blog.schoergendorfer.com/post/sempoi-rl-aligning-llm-semantic-reasoning-for-interpretable-out-of-town-poi-sequSecFeedde2026-09-01T08:17:50.098717+00:00SemPOI-RL: Aligning LLM Semantic Reasoning for Interpretable Out-of-Town POI Sequential Generationhttps://blog.schoergendorfer.com/post/using-grounded-theory-for-agent-behavior-analysis-at-scaleSecFeedde2026-09-01T08:17:50.041741+00:00Using Grounded Theory for Agent Behavior Analysis at Scalehttps://blog.schoergendorfer.com/post/dasc-decay-aware-state-compression-for-hybrid-linear-attention-servingSecFeedde2026-09-01T08:17:50.003420+00:00DASC: Decay-Aware State Compression for Hybrid Linear-Attention Servinghttps://blog.schoergendorfer.com/post/pave-predictive-alignment-and-value-guided-evolution-for-world-action-policiesSecFeedde2026-09-01T08:17:49.980599+00:00PAVE: Predictive Alignment and Value-Guided Evolution for World-Action Policieshttps://blog.schoergendorfer.com/post/beyond-ranking-accuracy-evaluating-llm-cited-feature-rationales-for-next-basketSecFeedde2026-09-01T08:17:49.957086+00:00Beyond Ranking Accuracy: Evaluating LLM-Cited Feature Rationales for Next Basket Repurchase Recommendationhttps://blog.schoergendorfer.com/post/beyond-token-level-guidance-inference-time-alignment-of-specialized-llms-via-croSecFeedde2026-09-01T08:17:49.915523+00:00Beyond Token-Level Guidance: Inference-Time Alignment of Specialized LLMs via Cross-Family Representation Steeringhttps://blog.schoergendorfer.com/post/online-estimation-of-dynamic-origin-destination-matrices-using-reinforcement-leaSecFeedde2026-09-01T08:17:49.895718+00:00Online Estimation of Dynamic Origin-Destination Matrices Using Reinforcement Learning with Link-Flow Propagation Guidancehttps://blog.schoergendorfer.com/post/one-ai-signal-many-human-judgments-a-bayesian-cascade-analysis-of-ai-based-crediSecFeedde2026-09-01T08:17:49.873679+00:00One AI Signal, Many Human Judgments: A Bayesian Cascade Analysis of AI-based Credibility Indicators in Online Information Spreadhttps://blog.schoergendorfer.com/post/tail-replay-escaping-the-curse-of-linear-attention-in-prefix-caching-for-hybridSecFeedde2026-09-01T08:17:49.849464+00:00Tail-Replay: Escaping the Curse of Linear Attention in Prefix Caching for Hybrid LLMshttps://blog.schoergendorfer.com/post/scenepilot-grow-and-repair-policy-for-text-driven-3d-indoor-scene-generationSecFeedde2026-09-01T08:17:49.826920+00:00ScenePilot: Grow-and-Repair Policy for Text-Driven 3D Indoor Scene Generationhttps://blog.schoergendorfer.com/post/catekv-on-sequential-consistency-for-long-context-llm-inference-accelerationSecFeedde2026-09-01T08:17:49.803098+00:00CateKV: On Sequential Consistency for Long-Context LLM Inference Accelerationhttps://blog.schoergendorfer.com/post/bcppo-bachelier-inspired-constrained-proximal-policy-optimization-for-tail-riskSecFeedde2026-09-01T08:17:49.781207+00:00BCPPO: Bachelier-Inspired Constrained Proximal Policy Optimization for Tail-Risk-Aware Safe Reinforcement Learninghttps://blog.schoergendorfer.com/post/centering-before-pruning-lightweight-geometry-correction-for-diversity-based-visSecFeedde2026-09-01T08:17:49.736407+00:00Centering before Pruning: Lightweight Geometry Correction for Diversity-Based Visual Token Pruning in LVLMshttps://blog.schoergendorfer.com/post/using-prosody-to-predict-syntactic-structureSecFeedde2026-09-01T08:17:49.712568+00:00Using Prosody to Predict Syntactic Structurehttps://blog.schoergendorfer.com/post/stratified-consistency-distillation-for-natural-language-formalizationSecFeedde2026-09-01T08:02:38.263023+00:00Stratified Consistency Distillation for Natural Language Formalizationhttps://blog.schoergendorfer.com/post/beyond-surface-forms-symbolic-edits-as-a-test-for-logical-reasoning-with-llmsSecFeedde2026-09-01T08:02:38.241231+00:00Beyond Surface Forms: Symbolic Edits as a Test for Logical Reasoning with LLMshttps://blog.schoergendorfer.com/post/motus2-a-self-evolving-general-world-model-for-dexterous-manipulationSecFeedde2026-09-01T08:02:38.216845+00:00Motus2: A Self-Evolving General World Model for Dexterous Manipulationhttps://blog.schoergendorfer.com/post/the-differential-reasoning-router-operationalizing-cost-aware-llm-annotation-inSecFeedde2026-09-01T08:02:38.188710+00:00The Differential Reasoning Router: Operationalizing Cost-Aware LLM Annotation in E-commercehttps://blog.schoergendorfer.com/post/sir-self-improving-red-teaming-for-compute-use-agentsSecFeedde2026-09-01T08:02:38.151835+00:00SIR: Self-improving Red-teaming for Compute Use Agentshttps://blog.schoergendorfer.com/post/science-sandboxes-measure-the-scientific-capability-of-ai-agentsSecFeedde2026-09-01T08:02:38.130157+00:00Science sandboxes measure the scientific capability of AI agentshttps://blog.schoergendorfer.com/post/cpr-for-llms-critical-point-routing-against-catastrophic-forgetting-in-domain-adSecFeedde2026-09-01T08:02:38.109060+00:00CPR for LLMs: Critical-Point Routing against Catastrophic Forgetting in Domain Adaptationhttps://blog.schoergendorfer.com/post/aligning-multi-trajectory-supervision-with-policy-optimization-for-vla-drivingSecFeedde2026-09-01T08:02:38.037823+00:00Aligning Multi-Trajectory Supervision with Policy Optimization for VLA Drivinghttps://blog.schoergendorfer.com/post/can-llms-take-the-pulse-of-the-economy-a-real-time-evaluation-of-llm-nowcasts-onSecFeedde2026-09-01T08:02:38.018786+00:00Can LLMs Take the Pulse of the Economy? A Real-Time Evaluation of LLM Nowcasts on Macroeconomic Indicatorshttps://blog.schoergendorfer.com/post/graph4bilo-graph-neural-network-approximation-for-bilevel-mixed-integer-linear-oSecFeedde2026-09-01T08:02:37.975374+00:00Graph4BiLO: Graph Neural Network Approximation for Bilevel Mixed-Integer Linear Optimizationhttps://blog.schoergendorfer.com/post/arkios-an-open-bilingual-english-nepali-language-model-trained-from-scratch-withSecFeedde2026-09-01T08:02:37.955687+00:00Arkios: An Open Bilingual English-Nepali Language Model Trained From Scratch, with a Devanagari-Aware Tokenizerhttps://blog.schoergendorfer.com/post/how-do-world-models-and-policies-compose-in-llm-agents-a-joint-spectral-and-behaSecFeedde2026-09-01T08:02:37.935467+00:00How do World Models and Policies Compose in LLM Agents? A Joint Spectral and Behavioral Accounthttps://blog.schoergendorfer.com/post/pak3h-evaluating-the-cost-of-cultural-mismatch-in-llm-alignment-with-a-human-conSecFeedde2026-09-01T08:02:37.911287+00:00Pak3H: Evaluating the Cost of Cultural Mismatch in LLM Alignment with a Human-Contextualized Urdu Benchmarkhttps://blog.schoergendorfer.com/post/reachability-based-capability-confinement-for-llm-agents-under-indirect-prompt-iSecFeedde2026-09-01T08:02:37.889225+00:00Reachability-Based Capability Confinement for LLM Agents under Indirect Prompt Injectionhttps://blog.schoergendorfer.com/post/forget-or-fine-tune-a-comparative-study-of-machine-unlearning-strategies-for-noiSecFeedde2026-09-01T08:02:37.859555+00:00Forget or Fine-tune? A Comparative Study of Machine Unlearning Strategies for Noisy Label Correctionhttps://blog.schoergendorfer.com/post/error-detection-for-petct-radiology-reports-domain-specific-vs-large-language-moSecFeedde2026-09-01T07:47:48.217486+00:00Error Detection for PET/CT Radiology Reports: Domain-Specific vs Large Language Modelshttps://blog.schoergendorfer.com/post/tempo-temporally-grounded-multi-task-post-training-for-large-audio-language-modeSecFeedde2026-09-01T07:47:48.196109+00:00TEMPO: Temporally-grounded Multi-task Post-training for Large Audio-Language Modelshttps://blog.schoergendorfer.com/post/partition-aware-unlearning-for-removing-spurious-correlations-in-large-vision-laSecFeedde2026-09-01T07:47:48.176455+00:00Partition-Aware Unlearning for Removing Spurious Correlations in Large Vision-Language Modelshttps://blog.schoergendorfer.com/post/generating-clinical-vignettes-that-preserve-cognitive-formulationsSecFeedde2026-09-01T07:47:48.155939+00:00Generating Clinical Vignettes that Preserve Cognitive Formulationshttps://blog.schoergendorfer.com/post/beyond-fluency-a-rubric-based-benchmark-for-evaluating-saudi-dialect-and-culturaSecFeedde2026-09-01T07:47:48.137057+00:00Beyond Fluency: A Rubric-Based Benchmark for Evaluating Saudi Dialect and Cultural Competence in Large Language Modelshttps://blog.schoergendorfer.com/post/training-free-action-correction-for-vla-model-failures-via-language-feedbackSecFeedde2026-09-01T07:47:48.117421+00:00Training-Free Action Correction for VLA Model Failures via Language Feedbackhttps://blog.schoergendorfer.com/post/the-policy-deficit-in-ai-x-social-emotional-learning-researchSecFeedde2026-09-01T07:47:48.100674+00:00The Policy Deficit in AI x Social-Emotional Learning Researchhttps://blog.schoergendorfer.com/post/influence-is-not-authority-when-causal-guardrail-signals-make-legitimate-tool-usSecFeedde2026-09-01T07:47:48.083606+00:00Influence Is Not Authority: When Causal Guardrail Signals Make Legitimate Tool Use Look Like an Attack in Tool-Using LLM Agentshttps://blog.schoergendorfer.com/post/hallucination-mitigation-for-large-vision-language-models-via-implicit-feature-sSecFeedde2026-09-01T07:47:48.064637+00:00Hallucination Mitigation for Large Vision-Language Models via Implicit Feature Stabilizationhttps://blog.schoergendorfer.com/post/indicdetect-evaluating-cross-lingual-llm-generated-text-detection-for-hindi-teluSecFeedde2026-09-01T07:47:48.030012+00:00IndicDetect: Evaluating Cross-Lingual LLM-Generated Text Detection for Hindi, Telugu, and Tamilhttps://blog.schoergendorfer.com/post/when-less-is-more-understanding-when-token-filtering-helps-and-fails-in-ai-generSecFeedde2026-09-01T07:47:48.004578+00:00When Less is More: Understanding When Token Filtering Helps and Fails in AI-generated Text Detectionhttps://blog.schoergendorfer.com/post/intervene-temporal-abstraction-interval-based-transformers-for-short-horizon-medSecFeedde2026-09-01T07:47:47.984045+00:00INTERVenE: Temporal-Abstraction-Interval Based Transformers for Short-Horizon Medical Event Predictionhttps://blog.schoergendorfer.com/post/reign-refurbished-embeddings-with-integrated-guidance-networks-for-efficient-conSecFeedde2026-09-01T07:47:47.958183+00:00REIGN: Refurbished Embeddings with Integrated Guidance Networks for Efficient Context-Length Scalinghttps://blog.schoergendorfer.com/post/r2a-learning-persona-policies-through-persona-representation-learning-and-runtimSecFeedde2026-09-01T07:47:47.938343+00:00R$^2$A: Learning Persona Policies Through Persona Representation Learning and Runtime Alignmenthttps://blog.schoergendorfer.com/post/syncrash-a-multi-stage-pipeline-for-zero-shot-accident-detection-and-localizatioSecFeedde2026-09-01T07:47:47.917360+00:00SynCrash: A Multi-Stage Pipeline for Zero-Shot Accident Detection and Localization in Traffic Surveillance Videohttps://blog.schoergendorfer.com/post/a-calibration-audit-of-confidence-in-feed-forward-3d-reconstructionSecFeedde2026-09-01T07:47:47.876661+00:00A Calibration Audit of Confidence in Feed-Forward 3D Reconstructionhttps://blog.schoergendorfer.com/post/medsegbenchmarker-a-raw-count-first-framework-for-controlled-2d-medical-image-seSecFeedde2026-09-01T07:47:47.849576+00:00MedSegBenchmarker: A Raw-Count-First Framework for Controlled 2D Medical Image Segmentation Benchmarkshttps://blog.schoergendorfer.com/post/cost-effective-repository-exploration-for-agentic-issue-localizationSecFeedde2026-09-01T07:32:48.083266+00:00Cost-Effective Repository Exploration for Agentic Issue Localizationhttps://blog.schoergendorfer.com/post/conducting-stylistic-analysis-of-paintings-through-an-art-history-agentSecFeedde2026-09-01T07:32:48.064921+00:00Conducting Stylistic Analysis of Paintings through an Art-History Agenthttps://blog.schoergendorfer.com/post/llmode-aligning-odes-with-llms-via-gated-token-injection-for-irregular-spatio-teSecFeedde2026-09-01T07:32:48.045500+00:00LLMODE: Aligning ODEs with LLMs via Gated Token Injection for Irregular Spatio-Temporal Forecastinghttps://blog.schoergendorfer.com/post/mi-distillation-selecting-from-model-interpolated-instruct-reasoning-data-spectrSecFeedde2026-09-01T07:32:48.028073+00:00MI-Distillation: Selecting from Model-Interpolated Instruct-Reasoning Data Spectrum for Chain-of-Thought Distillationhttps://blog.schoergendorfer.com/post/agenticrag-r1-agentic-reinforcement-learning-with-stack-memory-for-multi-step-reSecFeedde2026-09-01T07:32:48.008469+00:00AgenticRag-R1: Agentic Reinforcement Learning with Stack Memory for Multi-Step Reasoning, Retrieval and Memorizinghttps://blog.schoergendorfer.com/post/cineforge-self-improving-agents-for-long-horizon-video-generationSecFeedde2026-09-01T07:32:47.987412+00:00CineForge: Self-Improving Agents for Long-Horizon Video Generationhttps://blog.schoergendorfer.com/post/memory-first-fact-checking-a-knowledge-graph-grounded-multi-agent-system-for-misSecFeedde2026-09-01T07:32:47.969421+00:00Memory-First Fact-Checking: A Knowledge-Graph-Grounded Multi-Agent System for Misinformation Detectionhttps://blog.schoergendorfer.com/post/forward-deployed-full-stack-engineering-for-autonomous-cloud-mlopsSecFeedde2026-09-01T07:32:47.949621+00:00Forward-Deployed Full-Stack Engineering for Autonomous Cloud MLOpshttps://blog.schoergendorfer.com/post/wide-learning-learning-to-reach-evidenceSecFeedde2026-09-01T07:32:47.927681+00:00Wide Learning: Learning to Reach Evidencehttps://blog.schoergendorfer.com/post/sup-mimic-a-multi-task-clinical-diagnosis-benchmark-for-evaluating-llms-robustneSecFeedde2026-09-01T07:32:47.905195+00:00SUP-MIMIC: A Multi-Task Clinical Diagnosis Benchmark for Evaluating LLMs' Robustness to Contradictory Evidencehttps://blog.schoergendorfer.com/post/hoopmind-a-real-time-neural-game-tree-system-for-opponent-aware-possession-plannSecFeedde2026-09-01T07:32:47.884920+00:00HoopMind: A Real-Time Neural Game-Tree System for Opponent-Aware Possession Planninghttps://blog.schoergendorfer.com/post/physwave-physics-guided-latent-diffusion-models-for-controllable-spatial-audio-gSecFeedde2026-09-01T07:32:47.864152+00:00PhysWave: Physics-Guided Latent Diffusion Models for Controllable Spatial Audio Generationhttps://blog.schoergendorfer.com/post/evaluating-llms-on-conversational-text-to-sql-under-chain-ambiguity-and-intent-dSecFeedde2026-09-01T07:32:47.839875+00:00Evaluating LLMs on Conversational Text-to-SQL under Chain Ambiguity and Intent Drifthttps://blog.schoergendorfer.com/post/agm-achievement-grounded-memory-for-closed-loop-agents-with-frozen-vla-policiesSecFeedde2026-09-01T07:32:47.812818+00:00AGM: Achievement-Grounded Memory for Closed-Loop Agents with Frozen VLA Policieshttps://blog.schoergendorfer.com/post/integrating-adaptive-human-behavior-into-epidemic-models-with-large-language-modSecFeedde2026-09-01T07:32:47.790835+00:00Integrating adaptive human behavior into epidemic models with large language modelshttps://blog.schoergendorfer.com/post/the-emergent-symbolic-structure-of-artificial-neural-networksSecFeedde2026-09-01T07:32:47.769584+00:00The Emergent Symbolic Structure of Artificial Neural Networkshttps://blog.schoergendorfer.com/post/argument-aware-semantic-alignment-of-normative-texts-a-toulmin-based-neuro-symboSecFeedde2026-09-01T07:32:47.751054+00:00Argument-Aware Semantic Alignment of Normative Texts: A Toulmin-Based Neuro-Symbolic Approachhttps://blog.schoergendorfer.com/post/on-the-plasticity-collapse-in-continual-machine-unlearningSecFeedde2026-09-01T07:32:47.727851+00:00On the Plasticity Collapse in Continual Machine Unlearninghttps://blog.schoergendorfer.com/post/denoising-as-projection-constrained-optimization-with-gradient-guided-diffusionSecFeedde2026-09-01T07:32:47.701074+00:00Denoising as Projection: Constrained Optimization with Gradient-Guided Diffusionhttps://blog.schoergendorfer.com/post/generalizable-multi-agent-planning-from-signal-temporal-logic-specifications-viaSecFeedde2026-09-01T07:32:47.676418+00:00Generalizable Multi-Agent Planning from Signal Temporal Logic Specifications via Diffusionhttps://blog.schoergendorfer.com/post/applications-of-risk-science-to-ai-fairness-evaluation-principles-challenges-andSecFeedde2026-09-01T07:17:45.971166+00:00Applications of Risk Science to AI Fairness Evaluation: Principles, Challenges, and Best Practiceshttps://blog.schoergendorfer.com/post/muddle-measuring-understanding-of-documents-under-distractor-and-length-effectsSecFeedde2026-09-01T07:17:45.950995+00:00MUDDLE: Measuring Understanding of Documents under Distractor and Length Effectshttps://blog.schoergendorfer.com/post/knowledge-distillation-under-teacher-misspecification-an-order-parameter-analysiSecFeedde2026-09-01T07:17:45.932222+00:00Knowledge Distillation under Teacher Misspecification: An Order-Parameter Analysis of the Gap between Teacher Mimicry and Task Performancehttps://blog.schoergendorfer.com/post/chain-of-thought-faithfulness-of-reasoning-models-varies-with-where-and-how-prefSecFeedde2026-09-01T07:17:45.911335+00:00Chain-of-Thought Faithfulness of Reasoning Models Varies with Where and How Preference Cues Are Deliveredhttps://blog.schoergendorfer.com/post/benchmark-contamination-a-taxonomy-organized-by-defeated-mitigationSecFeedde2026-09-01T07:17:45.891026+00:00Benchmark Contamination: A Taxonomy Organized by Defeated Mitigationhttps://blog.schoergendorfer.com/post/reference-grafting-matches-fine-tuning-at-eliciting-sandbagged-capabilitiesSecFeedde2026-09-01T07:17:45.862362+00:00Reference-Grafting Matches Fine-Tuning at Eliciting Sandbagged Capabilitieshttps://blog.schoergendorfer.com/post/ai-can-be-easily-persuaded-in-clinical-decision-makingSecFeedde2026-09-01T07:17:45.840925+00:00AI Can Be Easily Persuaded in Clinical Decision Makinghttps://blog.schoergendorfer.com/post/ss-esoap-self-scaled-adaptive-preconditioning-for-physics-informed-learningSecFeedde2026-09-01T07:17:45.821832+00:00SS-ESOAP: Self-Scaled Adaptive Preconditioning for Physics-Informed Learninghttps://blog.schoergendorfer.com/post/does-latent-planning-survive-point-clouds-action-conditioned-jepa-world-models-fSecFeedde2026-09-01T07:17:45.802066+00:00Does Latent Planning Survive Point Clouds? Action-Conditioned JEPA World Models for Geometric Observationshttps://blog.schoergendorfer.com/post/polis-3d-self-supervision-at-city-scaleSecFeedde2026-09-01T07:17:45.780690+00:00Polis: 3D Self-Supervision at City Scalehttps://blog.schoergendorfer.com/post/scalable-clinical-data-infrastructure-and-comparative-ml-evaluation-for-hospitalSecFeedde2026-09-01T07:17:45.756224+00:00Scalable Clinical Data Infrastructure and Comparative ML Evaluation for Hospitalisation Risk Prediction in Elderly Patients with Multiple Long-Term Conditions using CPRDhttps://blog.schoergendorfer.com/post/safe-to-resume-breaking-execution-continuity-of-agent-execution-via-rollbackSecFeedde2026-09-01T07:17:45.709260+00:00Safe to Resume? Breaking Execution Continuity of Agent Execution via Rollbackhttps://blog.schoergendorfer.com/post/arabic-safety-alignment-as-selective-refusal-an-empirical-study-of-sft-dpo-and-gSecFeedde2026-09-01T07:17:45.686039+00:00Arabic Safety Alignment as Selective Refusal: An Empirical Study of SFT, DPO, and Guard Calibrationhttps://blog.schoergendorfer.com/post/stagewell-a-process-aligned-chinese-corpus-for-positive-psychology-support-dialoSecFeedde2026-09-01T07:17:45.664182+00:00StageWell: A Process-Aligned Chinese Corpus for Positive-Psychology Support Dialoguehttps://blog.schoergendorfer.com/post/sge-semantically-guided-exploration-for-unstructured-environments-via-image-spacSecFeedde2026-09-01T07:17:45.641602+00:00SGE: Semantically-Guided Exploration for Unstructured Environments via Image-Space Waypoint Samplinghttps://blog.schoergendorfer.com/post/hyper3-clip-hierarchy-conditioned-hyperbolic-vision-language-trainingSecFeedde2026-09-01T07:17:45.607116+00:00Hyper3-CLIP: Hierarchy-Conditioned Hyperbolic Vision-Language Traininghttps://blog.schoergendorfer.com/post/super-library-agent-joint-generation-and-maintenance-of-multiple-applications-beSecFeedde2026-09-01T07:17:45.584618+00:00Super Library Agent: Joint Generation and Maintenance of Multiple Applications Beyond the Single Codebasehttps://blog.schoergendorfer.com/post/evaluating-llm-based-ai-agents-integrated-with-materials-synthesis-tools-the-casSecFeedde2026-09-01T07:17:45.559248+00:00Evaluating LLM-based AI agents integrated with materials synthesis tools: the case of atomic layer depositionhttps://blog.schoergendorfer.com/post/improving-randomized-metric-distortion-to-23282SecFeedde2026-09-01T07:17:45.532991+00:00Improving Randomized Metric Distortion to 2.3282https://blog.schoergendorfer.com/post/detecting-and-repairing-hallucinations-in-retrieval-augmented-generationSecFeedde2026-09-01T07:03:19.983699+00:00Detecting and Repairing Hallucinations in Retrieval-Augmented Generationhttps://blog.schoergendorfer.com/post/learning-simple-test-time-environments-for-llm-web-agentsSecFeedde2026-09-01T07:03:19.957176+00:00Learning Simple Test-Time Environments for LLM Web Agentshttps://blog.schoergendorfer.com/post/when-do-larger-batches-help-scale-llm-reinforcement-learningSecFeedde2026-09-01T07:03:19.932229+00:00When Do Larger Batches Help Scale LLM Reinforcement Learning?https://blog.schoergendorfer.com/post/aoi-net-structural-face-aoi-guided-eye-gaze-track-representation-learning-for-auSecFeedde2026-09-01T07:03:19.909794+00:00AOI-Net: Structural Face AOI-Guided Eye-Gaze Track Representation Learning for Autism Spectrum Disorder Detectionhttps://blog.schoergendorfer.com/post/ragdiffusion-from-macro-retrieval-to-micro-fidelity-alignment-for-garment-generaSecFeedde2026-09-01T07:03:19.886684+00:00RAGDiffusion++: From Macro-Retrieval to Micro-Fidelity Alignment for Garment Generationhttps://blog.schoergendorfer.com/post/shadowbench-toward-reliable-automatic-evaluation-of-semantic-alignment-in-autofoSecFeedde2026-09-01T07:03:19.864604+00:00SHADOWBENCH: Toward Reliable Automatic Evaluation of Semantic Alignment in Autoformalizationhttps://blog.schoergendorfer.com/post/measurement-validity-in-llm-cultural-alignmentSecFeedde2026-09-01T07:03:19.842800+00:00Measurement Validity in LLM Cultural Alignmenthttps://blog.schoergendorfer.com/post/adaptive-multi-branching-for-shallow-decision-tree-inductionSecFeedde2026-09-01T07:03:19.819734+00:00Adaptive Multi-Branching for Shallow Decision Tree Inductionhttps://blog.schoergendorfer.com/post/agricam-a-track-mounted-crop-pollination-monitoring-robotSecFeedde2026-09-01T07:03:19.772616+00:00AGRICAM: A Track-Mounted Crop Pollination Monitoring Robothttps://blog.schoergendorfer.com/post/background-free-objectness-learning-for-class-agnostic-detectionSecFeedde2026-09-01T07:03:19.751020+00:00Background-Free Objectness Learning for Class-Agnostic Detectionhttps://blog.schoergendorfer.com/post/agentlogs-a-dataset-for-opening-the-black-box-of-githubs-cloud-agentSecFeedde2026-09-01T07:03:19.730192+00:00AgentLogs: A Dataset for Opening the Black Box of GitHub's Cloud Agenthttps://blog.schoergendorfer.com/post/pokaitrainer-scaling-belief-state-search-to-competitive-pokemon-vgcSecFeedde2026-09-01T07:03:19.709799+00:00PokaiTrainer: Scaling Belief-State Search to Competitive Pok\'emon VGChttps://blog.schoergendorfer.com/post/locked-at-the-entrance-open-inside-where-rlvr-narrows-the-solution-spaceSecFeedde2026-09-01T07:03:19.679984+00:00Locked at the Entrance, Open Inside: Where RLVR Narrows the Solution Spacehttps://blog.schoergendorfer.com/post/taal-mitigating-early-beam-pruning-in-generative-recommendation-via-temporal-autSecFeedde2026-09-01T07:03:19.660999+00:00TAAL: Mitigating Early Beam Pruning in Generative Recommendation via Temporal Autoregressive Alignmenthttps://blog.schoergendorfer.com/post/subtraction-based-tumor-segmentation-and-lesion-centered-pcr-prediction-for-theSecFeedde2026-09-01T07:03:19.640055+00:00Subtraction-Based Tumor Segmentation and Lesion-Centered pCR Prediction for the MAMA-MIA Challengehttps://blog.schoergendorfer.com/post/training-free-hidden-state-refinement-for-flow-matching-image-generatorsSecFeedde2026-09-01T07:03:19.619112+00:00Training-Free Hidden-State Refinement for Flow-Matching Image Generatorshttps://blog.schoergendorfer.com/post/starlinc-satellite-trail-artifact-removal-using-inter-frame-correlationSecFeedde2026-09-01T07:03:19.598622+00:00STARLINC: Satellite Trail Artifact Removal using Inter-Frame Correlationhttps://blog.schoergendorfer.com/post/not-the-same-protector-deployment-dependent-protective-intervention-in-llmsSecFeedde2026-09-01T07:03:19.570082+00:00Not the Same Protector: Deployment-Dependent Protective Intervention in LLMshttps://blog.schoergendorfer.com/post/cgfm-nav-cognitive-graph-field-memory-for-semantic-guided-lifelong-multimodal-emSecFeedde2026-09-01T06:47:45.668711+00:00CGFM-Nav: Cognitive Graph-Field Memory for Semantic-Guided Lifelong Multimodal Embodied Navigationhttps://blog.schoergendorfer.com/post/auditing-and-mitigating-privacy-leakage-in-cloud-edge-collaborative-decodingSecFeedde2026-09-01T06:47:45.649625+00:00Auditing and Mitigating Privacy Leakage in Cloud-Edge Collaborative Decodinghttps://blog.schoergendorfer.com/post/development-of-an-autonomous-ai-coding-agent-using-monte-carlo-tree-search-mctsSecFeedde2026-09-01T06:47:45.628147+00:00Development of an Autonomous AI Coding Agent using Monte Carlo Tree Search (MCTS) and Gemini LLM Frameworkshttps://blog.schoergendorfer.com/post/not-all-or-none-dynamic-construction-of-target-aware-memory-graph-for-conversatiSecFeedde2026-09-01T06:47:45.607900+00:00Not All or None: Dynamic Construction of Target-aware Memory Graph for Conversational Stance Detectionhttps://blog.schoergendorfer.com/post/docintent-answerability-guided-agentic-restoration-for-real-world-document-visuaSecFeedde2026-09-01T06:47:45.590043+00:00DocIntent: Answerability-Guided Agentic Restoration for Real-World Document Visual Question Answeringhttps://blog.schoergendorfer.com/post/a-unifying-perspective-on-language-model-representations-from-filler-role-structSecFeedde2026-09-01T06:47:45.572385+00:00A Unifying Perspective on Language Model Representations: From Filler-Role Structure to Mechanistic Interpretabilityhttps://blog.schoergendorfer.com/post/flow-jepa-flow-matching-for-robust-latent-dynamics-in-jepa-world-modelsSecFeedde2026-09-01T06:47:45.554307+00:00Flow-JEPA: Flow Matching for Robust Latent Dynamics in JEPA World Modelshttps://blog.schoergendorfer.com/post/rose-slam-robust-semantic-aware-gaussian-splatting-slam-from-dynamic-monocular-vSecFeedde2026-09-01T06:47:45.535207+00:00RoSe-SLAM: Robust Semantic-Aware Gaussian Splatting SLAM from Dynamic Monocular Videoshttps://blog.schoergendorfer.com/post/the-illusion-of-replacement-rethinking-specialized-machine-learning-models-in-thSecFeedde2026-09-01T06:47:45.517070+00:00The Illusion of Replacement: Rethinking Specialized Machine Learning Models in the Foundation Model Erahttps://blog.schoergendorfer.com/post/free-speech-and-artificial-intelligenceSecFeedde2026-09-01T06:47:45.497457+00:00Free Speech and Artificial Intelligencehttps://blog.schoergendorfer.com/post/the-information-geometry-of-product-reference-discrete-diffusion-interaction-groSecFeedde2026-09-01T06:47:45.479850+00:00The information geometry of product-reference discrete diffusion: Interaction growth complexity and optimal schedulinghttps://blog.schoergendorfer.com/post/diagnose-then-refine-a-closed-loop-tts-system-with-audiollm-guided-correctionSecFeedde2026-09-01T06:47:45.462482+00:00Diagnose, Then Refine: A Closed-Loop TTS System with AudioLLM-Guided Correctionhttps://blog.schoergendorfer.com/post/from-the-loss-landscape-to-diverse-feature-learning-in-neural-networksSecFeedde2026-09-01T06:47:45.439488+00:00From the Loss Landscape to Diverse Feature Learning in Neural Networkshttps://blog.schoergendorfer.com/post/activeaugment-online-active-learning-for-augmentation-selection-in-deep-learningSecFeedde2026-09-01T06:47:45.403711+00:00ActiveAugment: Online Active Learning for Augmentation Selection in Deep Learninghttps://blog.schoergendorfer.com/post/moving-the-mean-toward-the-known-good-not-beyond-it-what-inference-time-intervenSecFeedde2026-09-01T06:47:45.377614+00:00Moving the Mean Toward the Known Good, Not Beyond It: What Inference-Time Interventions and Weight Consolidation Buy in Open-Ended Generationhttps://blog.schoergendorfer.com/post/structured-state-reconciliation-for-human-ai-task-handoverSecFeedde2026-09-01T06:47:45.358731+00:00Structured State Reconciliation for Human-AI Task Handoverhttps://blog.schoergendorfer.com/post/hybrid-offline-online-multi-agent-decision-transformers-for-wireless-resource-maSecFeedde2026-09-01T06:47:45.331885+00:00Hybrid Offline-Online Multi-Agent Decision Transformers for Wireless Resource Managementhttps://blog.schoergendorfer.com/post/no-detectable-change-in-side-level-wer-from-prompt-level-context-a-preregisteredSecFeedde2026-09-01T06:47:45.303072+00:00No Detectable Change in Side-Level WER from Prompt-Level Context: A Preregistered Ablation on a Production Oral-History Corpushttps://blog.schoergendorfer.com/post/the-halt-vector-internalizing-a-causal-steering-intervention-for-efficient-reasoSecFeedde2026-09-01T06:32:48.378519+00:00The Halt Vector: Internalizing a Causal Steering Intervention for Efficient Reasoninghttps://blog.schoergendorfer.com/post/equivariant-sheaf-neural-networks-learning-geometric-transport-on-graphsSecFeedde2026-09-01T06:32:48.353974+00:00Equivariant Sheaf Neural Networks: Learning Geometric Transport on Graphshttps://blog.schoergendorfer.com/post/a-rigor-matched-audit-of-periodic-step-layer-skipping-for-efficient-llm-inferencSecFeedde2026-09-01T06:32:48.329010+00:00A rigor-matched audit of periodic-step layer skipping for efficient llm inference: conflayers versus swift, with a supplemental analysis of trained routing alternativeshttps://blog.schoergendorfer.com/post/toward-postural-state-classification-in-immersive-vr-with-multimodal-data-and-exSecFeedde2026-09-01T06:32:48.306192+00:00Toward Postural State Classification in Immersive VR with Multimodal Data and Explainability Analysishttps://blog.schoergendorfer.com/post/curvature-cryptanalysis-of-smooth-transformer-feed-forward-networksSecFeedde2026-09-01T06:32:48.284881+00:00Curvature Cryptanalysis of Smooth Transformer Feed-Forward Networkshttps://blog.schoergendorfer.com/post/delegating-before-learning-where-generative-ai-sits-in-students-professional-comSecFeedde2026-09-01T06:32:48.266153+00:00Delegating Before Learning: Where Generative AI Sits in Students' Professional Communicationhttps://blog.schoergendorfer.com/post/representation-learning-with-quantum-signal-processingSecFeedde2026-09-01T06:32:48.243546+00:00Representation Learning with Quantum Signal Processinghttps://blog.schoergendorfer.com/post/figmirror-ground-it-code-it-plot-itSecFeedde2026-09-01T06:32:48.208295+00:00FigMirror: Ground It, Code It, Plot Ithttps://blog.schoergendorfer.com/post/the-reach-of-a-verification-tool-decides-its-value-a-controlled-study-of-verificSecFeedde2026-09-01T06:32:48.190164+00:00The reach of a verification tool decides its value: A controlled study of verification surface, artifact quality, and cost in AI coding agentshttps://blog.schoergendorfer.com/post/a-large-scale-evaluation-of-text-guided-models-for-facial-editingSecFeedde2026-09-01T06:32:48.168448+00:00A Large-scale Evaluation of Text-guided Models for Facial Editinghttps://blog.schoergendorfer.com/post/astra-agentic-system-for-ticket-resolution-and-analysisSecFeedde2026-09-01T06:32:48.148824+00:00ASTRA - Agentic System for Ticket Resolution and Analysishttps://blog.schoergendorfer.com/post/robophys-3d-a-comprehensive-embodied-world-model-evaluation-via-3d-reconstructioSecFeedde2026-09-01T06:32:48.114049+00:00RoboPhys-3D: A Comprehensive Embodied World Model Evaluation via 3D Reconstructionhttps://blog.schoergendorfer.com/post/defending-wearable-vlms-against-private-attribute-inferenceSecFeedde2026-09-01T06:32:48.077592+00:00Defending Wearable VLMs Against Private Attribute Inferencehttps://blog.schoergendorfer.com/post/greenbench-benchmarking-energy-efficiency-and-carbon-footprint-of-open-source-llSecFeedde2026-09-01T06:32:48.055134+00:00GreenBench: Benchmarking Energy Efficiency and Carbon Footprint of Open-Source LLM Inference on Apple Siliconhttps://blog.schoergendorfer.com/post/test-time-scaling-for-scientific-equation-discoverySecFeedde2026-09-01T06:32:48.034645+00:00Test-Time Scaling for Scientific Equation Discoveryhttps://blog.schoergendorfer.com/post/can-large-language-models-identify-meaningful-touchpoints-in-conversion-attributSecFeedde2026-09-01T06:32:48.013862+00:00Can Large Language Models Identify Meaningful Touchpoints in Conversion Attribution?https://blog.schoergendorfer.com/post/cross-lingual-transfer-in-tulu-legal-comprehension-script-dependent-improvementSecFeedde2026-09-01T06:32:47.995169+00:00Cross Lingual Transfer in Tulu Legal Comprehension: Script-Dependent Improvement and RAG-Induced Knowledge Conflicthttps://blog.schoergendorfer.com/post/redesigning-and-auditing-deep-research-writing-for-faithful-reportsSecFeedde2026-09-01T06:17:47.932273+00:00Redesigning and Auditing Deep Research Writing for Faithful Reportshttps://blog.schoergendorfer.com/post/do-mllms-really-understand-low-resource-khmer-documents-a-pilot-study-on-khmer-dSecFeedde2026-09-01T06:17:47.884755+00:00Do MLLMs Really Understand Low-Resource Khmer Documents? A Pilot Study on Khmer Document VQAhttps://blog.schoergendorfer.com/post/pause-editable-strategy-artifacts-for-long-form-cultural-story-adaptationSecFeedde2026-09-01T06:17:47.867622+00:00PAUSE: Editable Strategy Artifacts for Long-Form Cultural Story Adaptationhttps://blog.schoergendorfer.com/post/enabling-proactive-spoken-turns-via-a-generalized-style-aware-full-duplex-framewSecFeedde2026-09-01T06:17:47.848621+00:00Enabling Proactive Spoken Turns via a Generalized Style-Aware Full-Duplex Frameworkhttps://blog.schoergendorfer.com/post/intelligent-identification-and-repair-of-design-defects-in-bim-via-domain-specifSecFeedde2026-09-01T06:17:47.828239+00:00Intelligent Identification and Repair of Design Defects in BIM via Domain-Specific Large Language Modelshttps://blog.schoergendorfer.com/post/do-large-language-models-scrutinise-what-they-review-a-multimodal-audit-of-scoriSecFeedde2026-09-01T06:17:47.803651+00:00Do large language models scrutinise what they review? A multimodal audit of scoring calibration, error detection, and author-identity effectshttps://blog.schoergendorfer.com/post/ma-rag-multi-agent-retrieval-augmented-generation-for-query-driven-summarizationSecFeedde2026-09-01T06:17:47.765817+00:00MA-RAG: Multi-Agent Retrieval-Augmented Generation for Query-Driven Summarization of Longitudinal Parkinson's Disease Assessmentshttps://blog.schoergendorfer.com/post/looking-again-measuring-sycophancy-in-the-reasoning-chains-of-multimodal-modelsSecFeedde2026-09-01T06:17:47.744917+00:00Looking Again: Measuring Sycophancy in the Reasoning Chains of Multimodal Models Under Pressurehttps://blog.schoergendorfer.com/post/puffer-incremental-fuzzy-deduplication-for-continuously-evolving-corporaSecFeedde2026-09-01T06:17:47.723077+00:00PUFFER: Incremental Fuzzy Deduplication for Continuously Evolving Corporahttps://blog.schoergendorfer.com/post/from-genai-virtual-patient-dialogue-logs-to-teacher-interpretable-process-evidenSecFeedde2026-09-01T06:17:47.702582+00:00From GenAI Virtual Patient Dialogue Logs to Teacher-Interpretable Process Evidence: A Learning Analytics Study in Higher Educationhttps://blog.schoergendorfer.com/post/stageet-stage-wise-typed-edit-tagging-for-grammatical-error-correction-with-arabSecFeedde2026-09-01T06:17:47.683938+00:00STAGEET: Stage-wise Typed Edit Tagging for Grammatical Error Correction with Arabic as a Case Studyhttps://blog.schoergendorfer.com/post/gurukul-ai-an-interactive-ai-driven-educational-platform-for-indian-education-sySecFeedde2026-09-01T06:17:47.663742+00:00Gurukul AI: An Interactive AI-Driven Educational Platform for Indian Education Systemhttps://blog.schoergendorfer.com/post/parametric-multimodal-user-memory-storing-what-captions-cannot-carrySecFeedde2026-09-01T06:17:47.645392+00:00Parametric Multimodal User Memory: Storing What Captions Cannot Carryhttps://blog.schoergendorfer.com/post/nlp-driven-knowledge-extraction-and-thematic-classification-of-translated-ancienSecFeedde2026-09-01T06:17:47.627442+00:00NLP-Driven Knowledge Extraction and Thematic Classification of Translated Ancient Indian Medical Textshttps://blog.schoergendorfer.com/post/powerslider-exploiting-phase-asymmetry-for-llm-serving-under-demand-responseSecFeedde2026-09-01T06:17:47.608294+00:00PowerSlider: Exploiting Phase Asymmetry for LLM Serving under Demand Responsehttps://blog.schoergendorfer.com/post/ontoaligner-ensemble-voting-based-fusion-across-heterogeneous-ontology-alignmentSecFeedde2026-09-01T06:17:47.561717+00:00OntoAligner-Ensemble: Voting-Based Fusion across Heterogeneous Ontology Alignment Techniqueshttps://blog.schoergendorfer.com/post/when-does-bigger-help-a-controlled-study-of-llm-scale-for-ontology-learningSecFeedde2026-09-01T06:02:32.031889+00:00When Does Bigger Help? A Controlled Study of LLM Scale for Ontology Learninghttps://blog.schoergendorfer.com/post/bloom-wilt-logit-tilting-for-behaviour-elicitation-in-automated-llm-auditingSecFeedde2026-09-01T06:02:32.009103+00:00BLOOM-WILT: Logit Tilting for Behaviour Elicitation in Automated LLM Auditinghttps://blog.schoergendorfer.com/post/cross-regional-grapevine-cold-hardiness-prediction-via-learned-multimodal-latentSecFeedde2026-09-01T06:02:31.983652+00:00Cross-Regional Grapevine Cold Hardiness Prediction via Learned Multimodal Latent Representationshttps://blog.schoergendorfer.com/post/token-efficient-data-reasoning-agents-via-adaptive-structuring-of-unstructured-dSecFeedde2026-09-01T06:02:31.958275+00:00Token-Efficient Data Reasoning Agents via Adaptive Structuring of Unstructured Datahttps://blog.schoergendorfer.com/post/reconciling-process-supervision-with-outcome-based-credit-in-agentic-policy-optiSecFeedde2026-09-01T06:02:31.933315+00:00Reconciling Process Supervision with Outcome-Based Credit in Agentic Policy Optimizationhttps://blog.schoergendorfer.com/post/scaling-large-reasoning-models-beyond-human-supervision-a-path-toward-superintelSecFeedde2026-09-01T06:02:31.907379+00:00Scaling Large Reasoning Models beyond Human Supervision: A Path toward Superintelligencehttps://blog.schoergendorfer.com/post/mnist-pro-mnist-is-back-as-a-partially-observable-world-for-ai-agentsSecFeedde2026-09-01T06:02:31.866088+00:00MNIST-PRO: MNIST is Back as a Partially Observable World for AI Agentshttps://blog.schoergendorfer.com/post/measure-before-you-manage-evaluating-agent-working-memory-in-coding-agentsSecFeedde2026-09-01T06:02:31.846325+00:00Measure Before You Manage: Evaluating Agent Working Memory in Coding Agentshttps://blog.schoergendorfer.com/post/learning-action-models-with-conditional-and-quantified-effects-via-uncertainty-gSecFeedde2026-09-01T06:02:31.824760+00:00Learning Action Models with Conditional and Quantified Effects via Uncertainty-Guided Explorationhttps://blog.schoergendorfer.com/post/carve-verified-expansion-for-variable-length-generation-in-diffusion-language-moSecFeedde2026-09-01T06:02:31.804907+00:00CARVE: Verified Expansion for Variable-Length Generation in Diffusion Language Modelshttps://blog.schoergendorfer.com/post/responsible-integration-of-ai-in-cancer-genomics-barriers-risks-and-pathways-toSecFeedde2026-09-01T06:02:31.775581+00:00Responsible Integration of AI in Cancer Genomics: Barriers, Risks, and Pathways to Trustworthy Clinical Translationhttps://blog.schoergendorfer.com/post/caer-causal-action-effect-reweighting-for-world-model-trainingSecFeedde2026-09-01T06:02:31.755983+00:00CAER: Causal Action Effect Reweighting for World Model Traininghttps://blog.schoergendorfer.com/post/predicting-residential-rents-in-dakar-using-machine-learningSecFeedde2026-09-01T06:02:31.734343+00:00Predicting Residential Rents in Dakar Using Machine Learninghttps://blog.schoergendorfer.com/post/vfr-audit-verdict-level-reliability-for-fairness-audits-in-hospital-length-of-stSecFeedde2026-09-01T06:02:31.708474+00:00VFR-Audit: Verdict-Level Reliability for Fairness Audits in Hospital Length-of-Stay Predictionhttps://blog.schoergendorfer.com/post/hsrm-hidden-state-reward-models-for-test-time-verificationSecFeedde2026-09-01T06:02:31.685356+00:00HSRM: Hidden-State Reward Models for Test-Time Verificationhttps://blog.schoergendorfer.com/post/skillzip-pro-execution-aware-dynamic-compression-of-progressively-loaded-skillsSecFeedde2026-09-01T06:02:31.659813+00:00SkillZip Pro: Execution-Aware Dynamic Compression of Progressively Loaded Skills for Self-Evolving Agentshttps://blog.schoergendorfer.com/post/which-rules-matter-now-policy-centroid-routing-before-an-intelligent-system-actsSecFeedde2026-09-01T06:02:31.634143+00:00Which Rules Matter Now? Policy-Centroid Routing Before an Intelligent System Actshttps://blog.schoergendorfer.com/post/autoregressive-mosaics-probing-2d-spatial-reasoning-in-text-only-language-modelsSecFeedde2026-09-01T06:02:31.611440+00:00Autoregressive Mosaics: Probing 2D Spatial Reasoning in Text-Only Language Modelshttps://blog.schoergendorfer.com/post/multimodal-adaptive-expert-selection-with-text-routing-and-ordinal-prototype-optSecFeedde2026-09-01T06:02:31.586569+00:00Multimodal Adaptive Expert Selection with Text Routing and Ordinal Prototype Optimization for Sentiment Analysishttps://blog.schoergendorfer.com/post/atlas-dual-horizon-diagnostic-evaluation-for-industrial-tool-use-agentsSecFeedde2026-09-01T05:47:28.991518+00:00ATLAS: Dual-Horizon Diagnostic Evaluation for Industrial Tool-Use Agentshttps://blog.schoergendorfer.com/post/medagent-r1-faithfulness-aware-reinforcement-learning-for-evidence-grounded-mediSecFeedde2026-09-01T05:47:28.959375+00:00MedAgent-R1: Faithfulness-Aware Reinforcement Learning for Evidence-Grounded Medical Reasoninghttps://blog.schoergendorfer.com/post/hirs-agent-a-hierarchical-multi-agent-system-for-reliable-long-horizon-remote-seSecFeedde2026-09-01T05:47:28.936682+00:00HiRS-Agent: A Hierarchical Multi-Agent System for Reliable Long-Horizon Remote Sensing Task Solvinghttps://blog.schoergendorfer.com/post/pykeen-nsx-a-modular-framework-for-static-dynamic-and-schema-aware-negative-sampSecFeedde2026-09-01T05:47:28.915058+00:00PyKEEN-NSX: A Modular Framework for Static, Dynamic and Schema-Aware Negative Sampling in PyKEENhttps://blog.schoergendorfer.com/post/geometry-of-divergence-tracking-hidden-state-trajectories-for-adaptive-multi-turSecFeedde2026-09-01T05:47:28.892569+00:00Geometry of Divergence: Tracking Hidden-State Trajectories for Adaptive Multi-Turn Reasoninghttps://blog.schoergendorfer.com/post/automated-testing-of-llm-based-post-hoc-explainers-using-model-checking-as-an-orSecFeedde2026-09-01T05:47:28.872587+00:00Automated Testing of LLM-Based Post Hoc Explainers Using Model Checking as an Oraclehttps://blog.schoergendorfer.com/post/turingllm-efficiently-scaling-foundation-models-toward-physical-aiSecFeedde2026-09-01T05:47:28.849477+00:00TuringLLM: Efficiently Scaling Foundation Models Toward Physical AIhttps://blog.schoergendorfer.com/post/adapath-query-adaptive-path-finding-via-path-bank-for-multi-hop-implicit-biomediSecFeedde2026-09-01T05:47:28.827873+00:00AdaPath: Query-Adaptive Path-Finding via Path-Bank for Multi-Hop Implicit Biomedical KGQAhttps://blog.schoergendorfer.com/post/garmentweaver-schema-aware-structured-synthesis-for-multimodal-sewing-patternsSecFeedde2026-09-01T05:47:28.803579+00:00GarmentWeaver: Schema-Aware Structured Synthesis for Multimodal Sewing Patternshttps://blog.schoergendorfer.com/post/designing-an-auditable-llm-supported-workflow-for-qualitative-thematic-analysisSecFeedde2026-09-01T05:47:28.781471+00:00Designing an Auditable LLM-Supported Workflow for Qualitative Thematic Analysishttps://blog.schoergendorfer.com/post/learning-assisted-congestion-aware-route-scheduling-for-semiconductor-fab-materiSecFeedde2026-09-01T05:47:28.742901+00:00Learning-Assisted Congestion-Aware Route Scheduling for Semiconductor Fab Material Control Systemshttps://blog.schoergendorfer.com/post/sciencearena-benchmarking-llms-on-latest-scientific-olympiad-competitionsSecFeedde2026-09-01T05:47:28.721681+00:00ScienceArena: Benchmarking LLMs on Latest Scientific Olympiad Competitionshttps://blog.schoergendorfer.com/post/chase-how-content-ecosystems-are-reshaped-when-ranking-is-the-only-targetSecFeedde2026-09-01T05:47:28.677235+00:00CHASE: How Content Ecosystems Are Reshaped When Ranking Is the Only Targethttps://blog.schoergendorfer.com/post/evoskill-injection-red-teaming-autonomous-skill-generation-and-evolution-in-selfSecFeedde2026-09-01T05:47:28.655414+00:00EvoSkill Injection: Red-Teaming Autonomous Skill Generation and Evolution in Self-Evolving Agentshttps://blog.schoergendorfer.com/post/from-metaheuristics-to-exact-methods-a-cp-sat-approach-for-multi-objective-healtSecFeedde2026-09-01T05:47:28.630677+00:00From Metaheuristics to Exact Methods: A CP-SAT Approach for Multi-Objective Healthcare Workforce Schedulinghttps://blog.schoergendorfer.com/post/derelab-probing-defeasible-reasoning-and-confirmation-bias-in-llms-with-a-generaSecFeedde2026-09-01T05:47:28.608499+00:00DERELAB: Probing Defeasible Reasoning and Confirmation Bias in LLMs with a Generative Benchmarkhttps://blog.schoergendorfer.com/post/dense-clinical-contrasts-enhance-medical-knowledge-updating-in-large-language-moSecFeedde2026-09-01T05:47:28.585895+00:00Dense Clinical Contrasts Enhance Medical Knowledge Updating in Large Language Modelshttps://blog.schoergendorfer.com/post/scaffolding-foundation-models-into-physical-world-agents-pushes-the-frontier-ofSecFeedde2026-09-01T05:47:28.563062+00:00Scaffolding Foundation Models into Physical-World Agents Pushes the Frontier of Long-Horizon Navigationhttps://blog.schoergendorfer.com/post/augmenting-human-performance-with-an-xr-agent-learning-from-online-behavior-andSecFeedde2026-09-01T05:32:30.175034+00:00Augmenting Human Performance with an XR Agent Learning from Online Behavior and BCI Evidencehttps://blog.schoergendorfer.com/post/will-the-user-ever-know-covert-indirect-prompt-injection-on-tool-using-llm-agentSecFeedde2026-09-01T05:32:30.152775+00:00Will the User Ever Know? Covert Indirect Prompt Injection on Tool-Using LLM Agentshttps://blog.schoergendorfer.com/post/co-annotator-expert-distilled-vit-and-vlm-for-visual-and-documentation-guidanceSecFeedde2026-09-01T05:32:30.132629+00:00Co-Annotator: Expert-Distilled ViT and VLM for Visual and Documentation Guidance in Age-Related Macular Degenerationhttps://blog.schoergendorfer.com/post/answer-probing-guided-search-for-diverse-solution-exploration-of-llmsSecFeedde2026-09-01T05:32:30.114921+00:00Answer Probing-Guided Search for Diverse Solution Exploration of LLMshttps://blog.schoergendorfer.com/post/ignorance-or-incompetence-constructing-knowledge-gated-verifiable-tasks-for-llmSecFeedde2026-09-01T05:32:30.095477+00:00Ignorance or Incompetence? Constructing Knowledge-Gated, Verifiable Tasks for LLM Agentshttps://blog.schoergendorfer.com/post/simcraft-distilling-remote-sensing-agents-via-synthetic-trajectories-and-contextSecFeedde2026-09-01T05:32:30.074028+00:00SimCRAFT: Distilling Remote Sensing Agents via Synthetic Trajectories and Contextual Retrieval-Augmented Fine-Tuninghttps://blog.schoergendorfer.com/post/generating-workflow-dags-from-natural-language-with-non-reasoning-llmsSecFeedde2026-09-01T05:32:30.051657+00:00Generating Workflow DAGs from Natural Language with Non-Reasoning LLMshttps://blog.schoergendorfer.com/post/rethinking-the-test-time-prompt-tuning-objective-from-the-perspective-of-calibraSecFeedde2026-09-01T05:32:29.978625+00:00Rethinking the Test-Time Prompt Tuning Objective from the Perspective of Calibrationhttps://blog.schoergendorfer.com/post/lamoc-loss-aware-modular-compression-for-llmsSecFeedde2026-09-01T05:32:29.960214+00:00LaMoC: Loss-Aware Modular Compression for LLMshttps://blog.schoergendorfer.com/post/spark-skeleton-guided-reasoning-synthesis-from-large-scale-scientific-literatureSecFeedde2026-09-01T05:32:29.939522+00:00SPARK: Skeleton-Guided Reasoning Synthesis from Large-Scale Scientific Literaturehttps://blog.schoergendorfer.com/post/favor-llm-based-agentic-framework-for-factor-mining-via-empirical-validationSecFeedde2026-09-01T05:32:29.919665+00:00FaVOR: LLM-Based Agentic Framework for Factor Mining via Empirical Validationhttps://blog.schoergendorfer.com/post/ax-k2-technical-reportSecFeedde2026-09-01T05:32:29.897695+00:00A.X K2 Technical Reporthttps://blog.schoergendorfer.com/post/vera-authority-preserving-edge-revocation-for-federated-ai-agent-workflowsSecFeedde2026-09-01T05:32:29.878499+00:00VERA: Authority-Preserving Edge Revocation for Federated AI-Agent Workflowshttps://blog.schoergendorfer.com/post/mitigating-over-optimization-in-prm-guided-search-in-mathematical-reasoning-by-oSecFeedde2026-09-01T05:32:29.858694+00:00Mitigating Over-Optimization in PRM-Guided Search in Mathematical Reasoning by Optimizing the Guidehttps://blog.schoergendorfer.com/post/spec2twin-chain-orchestrating-bi-level-optimization-with-llms-for-blockchain-digSecFeedde2026-09-01T05:32:29.820354+00:00Spec2Twin-Chain: Orchestrating Bi-Level Optimization with LLMs for Blockchain Digital Twin Constructionhttps://blog.schoergendorfer.com/post/can-llm-agents-discover-evaluating-creativity-on-ml-engineering-tasksSecFeedde2026-09-01T05:32:29.800257+00:00Can LLM Agents Discover? Evaluating Creativity on ML Engineering Taskshttps://blog.schoergendorfer.com/post/balance-of-benchmarks-semantic-density-reweighting-for-benchmark-multiplicity-anSecFeedde2026-09-01T05:32:29.780855+00:00Balance of Benchmarks: Semantic Density Reweighting for Benchmark Multiplicity and Task-Conditioned Evaluationhttps://blog.schoergendorfer.com/post/beyond-uncertainty-multi-solver-disagreement-rewards-for-self-evolving-reasoningSecFeedde2026-09-01T05:17:40.574922+00:00Beyond Uncertainty: Multi-Solver Disagreement Rewards for Self-Evolving Reasoning Curriculahttps://blog.schoergendorfer.com/post/interpreting-and-steering-for-safe-and-correct-code-generationSecFeedde2026-09-01T05:17:40.551490+00:00Interpreting and Steering for Safe and Correct Code Generationhttps://blog.schoergendorfer.com/post/automatic-conversion-of-nice-guidelines-to-an-executable-computational-model-usiSecFeedde2026-09-01T05:17:40.528802+00:00Automatic Conversion of NICE Guidelines to an Executable Computational Model Using Large Language Modelshttps://blog.schoergendorfer.com/post/autocrat-within-trajectory-joint-control-of-stochasticity-and-compute-for-llm-reSecFeedde2026-09-01T05:17:40.504144+00:00AutoCRAT: Within-trajectory Joint Control of Stochasticity and Compute for LLM Reasoninghttps://blog.schoergendorfer.com/post/an-open-source-event-driven-pipeline-for-cryptocurrency-market-data-ingestion-foSecFeedde2026-09-01T05:17:40.479644+00:00An Open-Source, Event-Driven Pipeline for Cryptocurrency Market Data: Ingestion, Forecasting, and On-Chain Fraud Detectionhttps://blog.schoergendorfer.com/post/edge-engine-for-deterministic-graph-evaluation-through-conversation-simulation-fSecFeedde2026-09-01T05:17:40.454140+00:00EDGE: Engine for Deterministic Graph Evaluation through Conversation Simulation from Graph Structured DSL Configurationhttps://blog.schoergendorfer.com/post/review-before-trust-source-grounded-integrity-gates-for-ai-assisted-personal-heaSecFeedde2026-09-01T05:17:40.422533+00:00Review Before Trust: Source-Grounded Integrity Gates for AI-Assisted Personal Health Recordshttps://blog.schoergendorfer.com/post/searchwiki-learning-to-build-and-navigate-knowledge-wikis-for-active-informationSecFeedde2026-09-01T05:17:40.395596+00:00SearchWiki: Learning to Build and Navigate Knowledge Wikis for Active Information Seekinghttps://blog.schoergendorfer.com/post/spatial-matryoshka-training-for-multi-granularity-visual-document-retrievalSecFeedde2026-09-01T05:17:40.360983+00:00Spatial Matryoshka Training for Multi-Granularity Visual Document Retrievalhttps://blog.schoergendorfer.com/post/acrosswam10a-modular-latent-world-action-stack-for-compact-robot-policiesSecFeedde2026-09-01T05:17:40.336709+00:00AcrossWAM1.0:A Modular Latent World-Action Stack for Compact Robot Policieshttps://blog.schoergendorfer.com/post/page-rag-provenance-aware-graph-evidence-promotion-for-fixed-budget-multi-hop-reSecFeedde2026-09-01T05:17:40.269262+00:00PAGE-RAG: Provenance-Aware Graph Evidence Promotion for Fixed-Budget Multi-hop Retrieval-Augmented Generationhttps://blog.schoergendorfer.com/post/frameworkers-a-dynamic-multi-agent-framework-for-ai-generated-video-productionSecFeedde2026-09-01T05:17:40.245391+00:00FRAMEWORKERS: A Dynamic Multi-Agent Framework for AI-Generated Video Productionhttps://blog.schoergendorfer.com/post/ideation-arena-evaluating-llm-generated-research-ideas-with-battle-style-human-eSecFeedde2026-09-01T05:17:40.225139+00:00Ideation Arena: Evaluating LLM Generated Research Ideas with Battle-style Human Expert Assessmenthttps://blog.schoergendorfer.com/post/detect-before-you-attribute-cascade-failure-attribution-for-multi-agent-systemsSecFeedde2026-09-01T05:17:40.202697+00:00Detect Before You Attribute: Cascade Failure Attribution for Multi-Agent Systemshttps://blog.schoergendorfer.com/post/llms-interpret-embeddings-organize-graphs-emerge-agent-driven-compilation-of-sciSecFeedde2026-09-01T05:17:40.178980+00:00LLMs Interpret, Embeddings Organize, Graphs Emerge: Agent-Driven Compilation of Scientific Knowledgehttps://blog.schoergendorfer.com/post/towards-a-systems-foundation-for-agentic-skills-architecture-lifecycle-and-securSecFeedde2026-09-01T05:17:40.155565+00:00Towards a Systems Foundation for Agentic Skills: Architecture, Lifecycle, and Securityhttps://blog.schoergendorfer.com/post/not-safe-for-all-auditing-the-dialect-penalty-in-text-to-image-safety-pipelinesSecFeedde2026-09-01T05:17:40.130798+00:00Not Safe for All: Auditing the Dialect Penalty in Text-to-Image Safety Pipelineshttps://blog.schoergendorfer.com/post/call-neighbours-yourself-graph-walks-with-destination-conditioned-on-policy-selfSecFeedde2026-09-01T05:17:40.098931+00:00Call Neighbours Yourself: Graph Walks with Destination-Conditioned On-Policy Self-Distillationhttps://blog.schoergendorfer.com/post/can-escalation-channels-redirect-reward-hacking-toward-defect-disclosureSecFeedde2026-09-01T05:02:47.066097+00:00Can escalation channels redirect reward hacking toward defect disclosure?https://blog.schoergendorfer.com/post/toward-latent-language-model-skills-steering-and-optimization-an-empirical-studySecFeedde2026-09-01T05:02:47.042105+00:00Toward Latent Language Model Skills Steering and Optimization: An Empirical Studyhttps://blog.schoergendorfer.com/post/evogenui-bench-evaluating-llms-as-multi-turn-generative-ui-assistantsSecFeedde2026-09-01T05:02:47.022020+00:00EvoGenUI-Bench: Evaluating LLMs as Multi-Turn Generative UI Assistantshttps://blog.schoergendorfer.com/post/evaluating-tiny-recursive-models-across-training-for-code-generationSecFeedde2026-09-01T05:02:46.997080+00:00Evaluating Tiny Recursive Models Across Training for Code Generationhttps://blog.schoergendorfer.com/post/foresight-9-prospective-and-process-aware-evaluation-of-adaptive-trading-agentsSecFeedde2026-09-01T05:02:46.949640+00:00FORESIGHT-9: Prospective and Process-Aware Evaluation of Adaptive Trading Agentshttps://blog.schoergendorfer.com/post/tracer-per-tool-context-retention-for-llm-agents-via-consequence-attributed-reinSecFeedde2026-09-01T05:02:46.921187+00:00TRACER: Per-Tool Context Retention for LLM Agents via Consequence-Attributed Reinforcement Learninghttps://blog.schoergendorfer.com/post/litesearch-vl-small-multimodal-search-agents-via-trajectory-distillation-and-synSecFeedde2026-09-01T05:02:46.884993+00:00LiteSearch-VL: Small Multimodal Search Agents via Trajectory Distillation and Synthetic Step-DPOhttps://blog.schoergendorfer.com/post/plant-inspired-ai-plants-as-inspiration-for-novel-problem-formulations-and-two-cSecFeedde2026-09-01T05:02:46.858151+00:00Plant-Inspired AI: Plants as Inspiration for Novel Problem Formulations, and Two Case Studieshttps://blog.schoergendorfer.com/post/appsolver-adaptive-patch-partitioning-for-point-wise-ship-flow-prediction-on-unsSecFeedde2026-09-01T05:02:46.829160+00:00APPSolver: Adaptive Patch Partitioning for Point-Wise Ship Flow Prediction on Unstructured Mesheshttps://blog.schoergendorfer.com/post/cross-relational-preference-learning-for-better-llm-instruction-followingSecFeedde2026-09-01T05:02:46.807060+00:00Cross-Relational Preference Learning for Better LLM Instruction Followinghttps://blog.schoergendorfer.com/post/extending-totalsegmentator-predicting-patient-and-acquisition-characteristics-frSecFeedde2026-09-01T05:02:46.785578+00:00Extending TotalSegmentator: Predicting Patient and Acquisition Characteristics from CT and MR Imageshttps://blog.schoergendorfer.com/post/bird-history-a-benchmark-for-history-driven-text-to-sql-with-fine-grained-knowleSecFeedde2026-09-01T05:02:46.762549+00:00BIRD-History: A Benchmark for History-Driven Text-to-SQL with Fine-Grained Knowledge Annotationshttps://blog.schoergendorfer.com/post/formal-concept-analysis-with-three-types-of-negationSecFeedde2026-09-01T05:02:46.741849+00:00Formal Concept Analysis with Three Types of Negationhttps://blog.schoergendorfer.com/post/predicting-future-organ-dysfunction-in-icu-patients-using-temporal-convolutionalSecFeedde2026-09-01T05:02:46.719716+00:00Predicting Future Organ Dysfunction in ICU Patients Using Temporal Convolutional Networks on MIMIC-IV Datahttps://blog.schoergendorfer.com/post/mmpcbench-benchmarking-multimodal-large-language-models-on-proactive-critique-ofSecFeedde2026-09-01T05:02:46.658387+00:00MMPCBench: Benchmarking Multimodal Large Language Models on Proactive Critique of Flawed Inputshttps://blog.schoergendorfer.com/post/epacache-error-propagation-aware-caching-for-accelerating-diffusion-based-visualSecFeedde2026-09-01T05:02:46.638512+00:00EpaCache: Error-Propagation-Aware Caching for Accelerating Diffusion-Based Visual Generationhttps://blog.schoergendorfer.com/post/racer-reinforced-agent-collaboration-for-explainable-reasoning-on-knowledge-grapSecFeedde2026-09-01T05:02:46.614887+00:00RACER: Reinforced Agent Collaboration for Explainable Reasoning on Knowledge Graphshttps://blog.schoergendorfer.com/post/guardianagent-policy-conditioned-risk-adaptive-anonymization-with-verified-adverSecFeedde2026-09-01T04:47:39.759703+00:00GuardianAgent: Policy-Conditioned Risk-Adaptive Anonymization with Verified Adversarial Escalationhttps://blog.schoergendorfer.com/post/validating-fkgin-soundness-assessment-in-llm-augmented-indian-food-knowledgeSecFeedde2026-09-01T04:47:39.736800+00:00Validating FKG.in: Soundness Assessment in LLM-Augmented Indian Food Knowledgehttps://blog.schoergendorfer.com/post/localizing-emergent-failures-in-agentic-ai-recovering-minimal-repair-families-viSecFeedde2026-09-01T04:47:39.712848+00:00Localizing Emergent Failures in Agentic AI: Recovering Minimal Repair Families via Counterfactual Replayhttps://blog.schoergendorfer.com/post/computational-depth-measurement-in-thermographic-video-overcoming-spatial-overfiSecFeedde2026-09-01T04:47:39.686081+00:00Computational Depth Measurement in Thermographic Video: Overcoming Spatial Overfitting via Spatio-Temporal Decouplinghttps://blog.schoergendorfer.com/post/imag-eval-a-language-grounded-framework-for-interpretable-text-to-image-instructSecFeedde2026-09-01T04:47:39.661830+00:00Imag-Eval: a language-grounded framework for interpretable Text-to-Image instruction following evaluationhttps://blog.schoergendorfer.com/post/hyper-fold-exploring-the-expressive-limit-of-sequence-geometry-learning-for-protSecFeedde2026-09-01T04:47:39.641476+00:00Hyper-Fold: Exploring the Expressive Limit of Sequence-Geometry Learning for Proteins via Hypergraph Modelinghttps://blog.schoergendorfer.com/post/benevolent-bias-in-multi-turn-human-agent-dialogueSecFeedde2026-09-01T04:47:39.616462+00:00Benevolent Bias in Multi-Turn Human-Agent Dialoguehttps://blog.schoergendorfer.com/post/how-identity-and-opinion-shape-political-sycophancy-in-llmsSecFeedde2026-09-01T04:47:39.597609+00:00How Identity and Opinion Shape Political Sycophancy in LLMshttps://blog.schoergendorfer.com/post/an-explainable-coherence-score-for-detecting-temporal-inconsistencies-in-politicSecFeedde2026-09-01T04:47:39.578632+00:00An Explainable Coherence Score for Detecting Temporal Inconsistencies in Political Newshttps://blog.schoergendorfer.com/post/judgepanel-a-compact-judge-with-panel-deliberation-via-adaptive-multi-reward-reiSecFeedde2026-09-01T04:47:39.558790+00:00JudgePanel: A Compact Judge with Panel Deliberation via Adaptive Multi-Reward Reinforcement Learninghttps://blog.schoergendorfer.com/post/more-perspectives-stronger-signals-multi-perspective-enhancement-and-progressiveSecFeedde2026-09-01T04:47:39.535650+00:00More Perspectives, Stronger Signals: Multi-Perspective Enhancement and Progressive Fusion for Multimodal Entity Representation Learninghttps://blog.schoergendorfer.com/post/apiflow-bench-measuring-whether-agents-survive-long-dependent-api-workflowsSecFeedde2026-09-01T04:47:39.514993+00:00APIFlow-Bench: Measuring Whether Agents Survive Long, Dependent API Workflowshttps://blog.schoergendorfer.com/post/emergent-misalignment-is-not-magicalSecFeedde2026-09-01T04:47:39.492743+00:00Emergent Misalignment Is Not Magicalhttps://blog.schoergendorfer.com/post/clustering-as-approximation-by-constrained-projectors-theory-and-guaranteesSecFeedde2026-09-01T04:47:39.455470+00:00Clustering as Approximation by Constrained Projectors: Theory and Guaranteeshttps://blog.schoergendorfer.com/post/safeatlas-vl-beyond-binary-multimodal-safety-with-large-scale-data-and-guard-modSecFeedde2026-09-01T04:47:39.431767+00:00SafeAtlas-VL: Beyond Binary Multimodal Safety with Large-Scale Data and Guard Modelshttps://blog.schoergendorfer.com/post/hania-planner-guided-multimodal-graph-evidence-selection-for-grounded-question-aSecFeedde2026-09-01T04:47:39.386088+00:00HANIA: Planner-Guided Multimodal Graph Evidence Selection for Grounded Question Answeringhttps://blog.schoergendorfer.com/post/nested-convex-body-chasing-for-online-optimization-with-evolving-feasible-setsSecFeedde2026-09-01T04:47:39.365784+00:00Nested Convex-Body Chasing for Online Optimization with Evolving Feasible Setshttps://blog.schoergendorfer.com/post/revolutionizing-turn-by-turn-navigation-with-cloud-edge-deep-learningSecFeedde2026-09-01T04:32:48.857337+00:00Revolutionizing Turn-by-Turn Navigation with Cloud-Edge Deep Learninghttps://blog.schoergendorfer.com/post/let-prompts-bridge-defense-knowledge-transferable-graph-purification-via-vulneraSecFeedde2026-09-01T04:32:48.812797+00:00Let Prompts Bridge Defense Knowledge: Transferable Graph Purification via Vulnerability-Aware GPLhttps://blog.schoergendorfer.com/post/emolasp-emotion-recognition-with-language-models-and-answer-set-programmingSecFeedde2026-09-01T04:32:48.792047+00:00EmoLASP: Emotion Recognition with Language Models and Answer Set Programminghttps://blog.schoergendorfer.com/post/learning-to-follow-in-context-watermark-instructions-via-self-distillationSecFeedde2026-09-01T04:32:48.772447+00:00Learning to Follow In-Context Watermark Instructions via Self-Distillationhttps://blog.schoergendorfer.com/post/facts-without-rules-boundary-metadata-collapse-in-multi-agent-llm-handoffsSecFeedde2026-09-01T04:32:48.752109+00:00Facts Without Rules: Boundary Metadata Collapse in Multi-Agent LLM Handoffshttps://blog.schoergendorfer.com/post/disentangling-representation-using-attributes-based-gaussian-estimation-for-mediSecFeedde2026-09-01T04:32:48.733100+00:00Disentangling Representation using Attributes-based Gaussian Estimation for Medical Sound Diagnosishttps://blog.schoergendorfer.com/post/frequency-selective-neural-networks-as-a-foundation-architecture-for-time-seriesSecFeedde2026-09-01T04:32:48.710933+00:00Frequency Selective Neural Networks as a Foundation Architecture for Time Series Learninghttps://blog.schoergendorfer.com/post/multi-step-forecasting-of-grape-berry-temperature-based-on-lstm-model-with-feedSecFeedde2026-09-01T04:32:48.686929+00:00Multi-Step Forecasting of Grape Berry Temperature based on LSTM Model with Feed-Forward Attentionhttps://blog.schoergendorfer.com/post/verification-abundance-adjudication-scarcity-what-happens-to-mathematical-knowleSecFeedde2026-09-01T04:32:48.663437+00:00Verification abundance, adjudication scarcity: what happens to mathematical knowledge when proof checking becomes freehttps://blog.schoergendorfer.com/post/agentic-ai-uncovers-conserved-cross-tissue-protein-co-abundance-programs-inaccesSecFeedde2026-09-01T04:32:48.641844+00:00Agentic AI uncovers conserved cross-tissue protein co-abundance programs inaccessible to single-dataset analysishttps://blog.schoergendorfer.com/post/selective-forgetting-a-graph-based-memory-framework-for-long-term-llm-agentsSecFeedde2026-09-01T04:32:48.601247+00:00Selective Forgetting: A Graph-Based Memory Framework for Long-Term LLM Agentshttps://blog.schoergendorfer.com/post/the-role-of-network-topology-and-opponent-information-in-shaping-cooperation-inSecFeedde2026-09-01T04:32:48.577249+00:00The Role of Network Topology and Opponent Information in Shaping Cooperation in Multi-Agent Reinforcement Learning Systemshttps://blog.schoergendorfer.com/post/from-analytics-to-tumor-boards-an-evidence-linked-multi-agent-workflow-for-oncolSecFeedde2026-09-01T04:32:48.554566+00:00From Analytics to Tumor Boards: An Evidence-Linked Multi-Agent Workflow for Oncology Feature Extractionhttps://blog.schoergendorfer.com/post/efficient-gpu-retrieval-for-semantic-searchSecFeedde2026-09-01T04:32:48.525222+00:00Efficient GPU Retrieval for Semantic Searchhttps://blog.schoergendorfer.com/post/from-location-phrases-to-geographic-entities-task-adapted-retrieval-for-people-sSecFeedde2026-09-01T04:32:48.500198+00:00From Location Phrases to Geographic Entities: Task-Adapted Retrieval for People Searchhttps://blog.schoergendorfer.com/post/automated-researchers-can-reliably-mitigate-alignment-failuresSecFeedde2026-09-01T04:32:48.466300+00:00Automated Researchers Can Reliably Mitigate Alignment Failureshttps://blog.schoergendorfer.com/post/oculi-a-conversational-agentic-platform-for-automated-credit-risk-analysisSecFeedde2026-09-01T04:32:48.438842+00:00Oculi: A Conversational Agentic Platform for Automated Credit Risk Analysishttps://blog.schoergendorfer.com/post/evaluating-the-hidden-costs-of-personalization-in-large-language-modelsSecFeedde2026-09-01T04:32:48.277095+00:00Evaluating the Hidden Costs of Personalization in Large Language Modelshttps://blog.schoergendorfer.com/post/discovering-machine-correlates-of-consciousnessSecFeedde2026-09-01T04:17:41.188130+00:00Discovering Machine Correlates of Consciousnesshttps://blog.schoergendorfer.com/post/explainable-artificial-intelligence-xai-in-computational-pathology-definitions-tSecFeedde2026-09-01T04:17:41.166628+00:00Explainable Artificial Intelligence (XAI) in Computational Pathology: Definitions, Taxonomy, and Recommendationshttps://blog.schoergendorfer.com/post/capability-stratified-degradation-in-ternary-language-modelsSecFeedde2026-09-01T04:17:41.142658+00:00Capability-Stratified Degradation in Ternary Language Modelshttps://blog.schoergendorfer.com/post/efficient-geothermal-well-control-optimization-via-diffusion-surrogate-reinforceSecFeedde2026-09-01T04:17:41.101611+00:00Efficient Geothermal Well-Control Optimization via Diffusion-Surrogate Reinforcement Learninghttps://blog.schoergendorfer.com/post/permitgpt-a-unified-generative-ai-pipeline-for-construction-hazard-forecasting-pSecFeedde2026-09-01T04:17:41.079156+00:00PermitGPT: A Unified Generative-AI Pipeline for Construction Hazard Forecasting, Permit Prediction, and Community Impacthttps://blog.schoergendorfer.com/post/beyond-the-answer-key-robustness-evaluation-of-large-language-models-for-step-leSecFeedde2026-09-01T04:17:41.056302+00:00Beyond the Answer Key: Robustness Evaluation of Large Language Models for Step-Level Mathematical Verificationhttps://blog.schoergendorfer.com/post/pro-router-token-aware-progressive-model-routing-with-adaptive-edge-cloud-collabSecFeedde2026-09-01T04:17:41.030560+00:00Pro-Router: Token-Aware Progressive Model Routing with Adaptive Edge-Cloud Collaboration for Efficient Multimodal LLM Inferencehttps://blog.schoergendorfer.com/post/frac-mas-a-safe-and-explainable-multi-agent-system-for-fracture-diagnosisSecFeedde2026-09-01T04:17:40.986110+00:00FRAC-MAS: A Safe and Explainable Multi-Agent System for Fracture Diagnosishttps://blog.schoergendorfer.com/post/how-language-models-choose-sides-internal-representations-of-instruction-hierarcSecFeedde2026-09-01T04:17:40.942616+00:00How Language Models Choose Sides: Internal Representations of Instruction Hierarchyhttps://blog.schoergendorfer.com/post/self-specialized-teachers-for-domain-post-trainingSecFeedde2026-09-01T04:17:40.920750+00:00Self-Specialized Teachers for Domain Post-Traininghttps://blog.schoergendorfer.com/post/biasmix-finance-post-generation-kyc-guardrails-for-llm-portfolio-adviceSecFeedde2026-09-01T04:17:40.900184+00:00BiasMix-Finance: Post-Generation KYC Guardrails for LLM Portfolio Advicehttps://blog.schoergendorfer.com/post/from-extraction-to-governed-memory-multi-agent-knowledge-graph-construction-withSecFeedde2026-09-01T04:17:40.874223+00:00From Extraction to Governed Memory: Multi-Agent Knowledge Graph Construction with Domain-Expert Reviewhttps://blog.schoergendorfer.com/post/reward-oracle-mcts-for-formal-theorem-proving-sample-efficient-search-and-the-neSecFeedde2026-09-01T04:17:40.852066+00:00Reward-Oracle MCTS for Formal Theorem Proving: Sample-Efficient Search and the Need for Kernel-Level Proof Auditinghttps://blog.schoergendorfer.com/post/ai-scientist-mission-control-aimc-visual-analytics-for-human-oversight-of-autonoSecFeedde2026-09-01T04:17:40.827017+00:00AI Scientist Mission Control (AIMC): Visual Analytics for Human Oversight of Autonomous Scientific Discoveryhttps://blog.schoergendorfer.com/post/self-evolving-skills-via-surrogate-guided-solve-and-reproduceSecFeedde2026-09-01T04:17:40.804083+00:00Self-Evolving Skills via Surrogate-Guided Solve-and-Reproducehttps://blog.schoergendorfer.com/post/autoscientist-quant-self-evolving-coding-agents-for-automatic-research-in-quantiSecFeedde2026-09-01T04:17:40.777642+00:00AutoScientist-Quant: Self-Evolving Coding Agents for Automatic Research in Quantitative Investmenthttps://blog.schoergendorfer.com/post/crossaudit-a-git-native-cross-vendor-audit-loop-for-agentic-scienceSecFeedde2026-09-01T04:17:40.751833+00:00CrossAudit: A Git-Native, Cross-Vendor Audit Loop for Agentic Sciencehttps://blog.schoergendorfer.com/post/cdep-agent-connecting-meteorologically-detected-temporal-compound-events-to-realSecFeedde2026-09-01T04:02:52.734626+00:00CDEP Agent: Connecting Meteorologically Detected Temporal Compound Events to Real-World Documentary Evidencehttps://blog.schoergendorfer.com/post/machine-learning-enhanced-tabu-search-for-tactical-wireless-network-designSecFeedde2026-09-01T04:02:52.714285+00:00Machine Learning-Enhanced Tabu Search for Tactical Wireless Network Designhttps://blog.schoergendorfer.com/post/preference-elicitation-for-policy-optimization-and-application-to-aligning-heartSecFeedde2026-09-01T04:02:52.692246+00:00Preference Elicitation for Policy Optimization and Application to Aligning Heart Transplantation with Human Valueshttps://blog.schoergendorfer.com/post/regdivergence-101-an-llm-benchmark-for-cross-jurisdiction-regulatory-contradictiSecFeedde2026-09-01T04:02:52.636871+00:00RegDivergence-101: An LLM Benchmark for Cross-Jurisdiction Regulatory Contradiction Detection in Life Scienceshttps://blog.schoergendorfer.com/post/medtvl-harnessing-vision-and-language-for-medical-time-series-classificationSecFeedde2026-09-01T04:02:52.618570+00:00MedTVL: Harnessing Vision and Language for Medical Time Series Classificationhttps://blog.schoergendorfer.com/post/c3-unimm-causal-cycle-consistent-unified-multimodal-modeling-via-super-alignmentSecFeedde2026-09-01T04:02:52.601176+00:00C3-UniMM: Causal Cycle-Consistent Unified Multimodal Modeling via Super Alignment and Shared Decoding Spacehttps://blog.schoergendorfer.com/post/integrating-triaxial-imu-sensors-and-ensemble-learning-for-effective-parkinson-dSecFeedde2026-09-01T04:02:52.582222+00:00Integrating Triaxial IMU Sensors and Ensemble Learning for Effective Parkinson Disease Severity Classificationhttps://blog.schoergendorfer.com/post/leveraging-generative-ai-to-design-accessible-interactive-visualizations-for-undSecFeedde2026-09-01T04:02:52.563872+00:00Leveraging Generative AI to Design Accessible Interactive Visualizations for Undergraduate Mathematics: A Six-Phase Workflowhttps://blog.schoergendorfer.com/post/shape-of-chain-of-thought-in-math-reasoningSecFeedde2026-09-01T04:02:52.544788+00:00SHAPE of Chain-of-Thought in Math Reasoninghttps://blog.schoergendorfer.com/post/cdpr-counterfactual-advantage-based-credit-assignment-for-cost-aware-sequentialSecFeedde2026-09-01T04:02:52.524441+00:00CDPR: Counterfactual Advantage-based Credit Assignment for Cost-Aware Sequential Medical Diagnosishttps://blog.schoergendorfer.com/post/the-race-between-agentic-ai-capabilities-and-data-quality-control-in-online-survSecFeedde2026-09-01T04:02:52.503502+00:00The Race between Agentic AI Capabilities and Data Quality Control in Online Surveyshttps://blog.schoergendorfer.com/post/paper-pilot-a-human-in-the-loop-expert-system-for-evidence-traceable-scientificSecFeedde2026-09-01T04:02:52.482999+00:00Paper Pilot: A Human-in-the-Loop Expert System for Evidence-Traceable Scientific Manuscript Generation in Applied Scienceshttps://blog.schoergendorfer.com/post/the-signal-in-the-noise-an-auditable-reliability-layer-for-biomedical-text-classSecFeedde2026-09-01T04:02:52.463178+00:00The Signal in the Noise: An Auditable Reliability Layer for Biomedical Text Classificationhttps://blog.schoergendorfer.com/post/from-question-first-to-analyst-first-domain-expert-skills-and-verified-knowledgeSecFeedde2026-09-01T04:02:52.444770+00:00From Question-First to Analyst-First: Domain-Expert Skills and Verified Knowledge Compilation for Proactive Enterprise Analyticshttps://blog.schoergendorfer.com/post/statutory-ai-aligning-large-language-models-with-legal-normsSecFeedde2026-09-01T04:02:52.420703+00:00Statutory AI: Aligning Large Language Models With Legal Normshttps://blog.schoergendorfer.com/post/a-collective-capability-boundary-in-frontier-large-language-models-on-guidelineSecFeedde2026-09-01T04:02:52.402561+00:00A collective capability boundary in frontier large language models on guideline-conformant and case-specific oncology decision-makinghttps://blog.schoergendorfer.com/post/expert-validated-stem-qaSecFeedde2026-09-01T04:02:52.383118+00:00Expert-validated STEM QAhttps://blog.schoergendorfer.com/post/ds-lighting-making-agent-harnesses-explicit-for-data-science-automationSecFeedde2026-09-01T04:02:52.362451+00:00DS-Lighting: Making Agent Harnesses Explicit for Data-Science Automationhttps://blog.schoergendorfer.com/post/gradium-ai-releases-new-default-tts-model-810-hard-case-pass-rate-at-216-ms-timeSecFeedde2026-09-01T03:47:27.853493+00:00Gradium AI Releases New Default TTS Model: 81.0% Hard-Case Pass Rate at 216 ms Time-to-First-Audiohttps://blog.schoergendorfer.com/post/financially-motivated-threat-actor-breeze-comet-targets-brazilSecFeedde2026-09-01T03:17:33.865337+00:00Financially Motivated Threat Actor BREEZE COMET Targets Brazilhttps://blog.schoergendorfer.com/post/guildma-astaroth-malware-infection-from-brazilian-portuguese-email-tue-sep-1stSecFeedde2026-09-01T00:47:47.228297+00:00Guildma (Astaroth) malware infection from Brazilian Portuguese email, (Tue, Sep 1st)https://blog.schoergendorfer.com/post/introducing-wraptureSecFeedde2026-09-01T00:17:25.730168+00:00Introducing wrapturehttps://blog.schoergendorfer.com/post/kubernetes-v137-storage-version-migration-enabled-by-defaultSecFeedde2026-08-31T23:47:29.110475+00:00Kubernetes v1.37: Storage Version Migration Enabled by Defaulthttps://blog.schoergendorfer.com/post/keenable-ai-open-sources-needle-a-live-search-benchmark-that-rebuilds-its-querySecFeedde2026-08-31T23:47:29.076747+00:00Keenable AI Open-Sources NEEDLE: A Live Search Benchmark That Rebuilds Its Query Set Every Hourhttps://blog.schoergendorfer.com/post/the-hugging-face-hack-could-indicate-cultural-issues-at-openaiSecFeedde2026-08-31T23:32:42.647066+00:00The Hugging Face hack could indicate cultural issues at OpenAIhttps://blog.schoergendorfer.com/post/amazon-kinesis-data-streams-announces-streaming-tables-delivering-data-to-apacheSecFeedde2026-08-31T23:17:33.092675+00:00Amazon Kinesis Data Streams announces streaming tables, delivering data to Apache Iceberg tables on Amazon S3 Tableshttps://blog.schoergendorfer.com/post/amazon-kinesis-data-streams-announces-data-delivery-to-general-purpose-amazon-s3SecFeedde2026-08-31T23:17:33.069188+00:00Amazon Kinesis Data Streams announces data delivery to general purpose Amazon S3 bucketshttps://blog.schoergendorfer.com/post/amazon-workspaces-applications-is-now-available-in-three-new-aws-regionsSecFeedde2026-08-31T23:17:33.037611+00:00Amazon WorkSpaces Applications is now available in three new AWS Regions