https://blog.schoergendorfer.com/post/anthropics-claude-fixed-all-10-alignment-failures-then-it-tried-to-cheat-24-of-tSecFeedde2026-08-31T13:33:22.196476+00:00Anthropic’s Claude fixed all 10 alignment failures. Then it tried to cheat 2.4% of the time.https://blog.schoergendorfer.com/post/broadcom-launches-trusted-artifact-service-for-spring-frameworkSecFeedde2026-08-31T13:33:22.163009+00:00Broadcom Launches Trusted Artifact Service for Spring Frameworkhttps://blog.schoergendorfer.com/post/introducing-adaptive-intelligence-undermining-the-economics-of-every-bot-attackSecFeedde2026-08-31T13:33:22.122985+00:00Introducing Adaptive Intelligence: undermining the economics of every bot attackhttps://blog.schoergendorfer.com/post/usn-8699-1-libssh-vulnerabilitiesSecFeedde2026-08-31T13:17:43.411571+00:00USN-8699-1: libssh vulnerabilitieshttps://blog.schoergendorfer.com/post/31th-august-threat-intelligence-reportSecFeedde2026-08-31T13:17:42.438900+00:0031th August – Threat Intelligence Reporthttps://blog.schoergendorfer.com/post/your-agent-context-needs-a-development-lifecycleSecFeedde2026-08-31T13:02:59.976781+00:00Your agent context needs a development lifecyclehttps://blog.schoergendorfer.com/post/secure-by-default-is-your-only-way-forwardSecFeedde2026-08-31T13:02:59.942347+00:00Secure by default is your only way forwardhttps://blog.schoergendorfer.com/post/deepseeks-first-vision-model-vs-gemini-37-flash-it-comes-down-to-spend-vs-speedSecFeedde2026-08-31T12:02:52.010112+00:00DeepSeek’s first vision model vs. Gemini 3.7 Flash: It comes down to spend vs. speedhttps://blog.schoergendorfer.com/post/opentelemetry-has-graduated-now-whatSecFeedde2026-08-31T11:32:43.723610+00:00OpenTelemetry has graduated… now what?https://blog.schoergendorfer.com/post/tide-launched-raziel-for-ai-security-it-assumes-hackers-are-insideSecFeedde2026-08-31T10:17:50.714477+00:00Tide launched Raziel for AI security. It assumes hackers are inside.https://blog.schoergendorfer.com/post/vs-code-1135-gives-ai-coding-agents-a-second-opinionSecFeedde2026-08-31T10:17:50.686294+00:00VS Code 1.135 Gives AI Coding Agents a Second Opinionhttps://blog.schoergendorfer.com/post/valleyrat-masquerading-as-adwareSecFeedde2026-08-31T10:17:50.607813+00:00ValleyRAT masquerading as adwarehttps://blog.schoergendorfer.com/post/a-simple-website-summary-just-exposed-the-limits-of-ai-coding-guardrailsSecFeedde2026-08-31T09:47:35.535383+00:00A Simple Website Summary Just Exposed the Limits of AI Coding Guardrailshttps://blog.schoergendorfer.com/post/optimize-eks-operations-with-agents-reduce-mttr-with-aws-devops-agent-and-a-kubeSecFeedde2026-08-31T09:33:23.630006+00:00Optimize EKS operations with agents: Reduce MTTR with AWS DevOps Agent and a Kubernetes Operatorhttps://blog.schoergendorfer.com/post/github-tightens-copilots-billing-and-governance-rules-ahead-of-a-busy-fallSecFeedde2026-08-31T09:18:31.264691+00:00GitHub Tightens Copilot’s Billing and Governance Rules Ahead of a Busy Fallhttps://blog.schoergendorfer.com/post/lwiai-podcast-255-gemini-37-jalapeno-qwen-38-dronesSecFeedde2026-08-31T08:47:29.805886+00:00LWiAI Podcast #255 - Gemini 3.7, Jalapeño, Qwen 3.8, Droneshttps://blog.schoergendorfer.com/post/usn-8689-1-openjdk-26-vulnerabilitiesSecFeedde2026-08-31T08:02:45.679100+00:00USN-8689-1: OpenJDK 26 vulnerabilitieshttps://blog.schoergendorfer.com/post/openclaw-releases-openclaw-20-guided-model-setup-575-ms-control-ui-startup-and-oSecFeedde2026-08-31T08:02:45.653190+00:00OpenClaw Releases OpenClaw 2.0: Guided Model Setup, 575 ms Control UI Startup, and One Trust Boundary Per Gatewayhttps://blog.schoergendorfer.com/post/when-stale-constraints-go-unchecked-budgeted-verification-failures-in-inherited-3SecFeedde2026-08-31T08:02:45.629872+00:00When Stale Constraints Go Unchecked: Budgeted Verification Failures in Inherited Agent Memoryhttps://blog.schoergendorfer.com/post/set-shifting-behavioral-test-for-harnessed-agents-2SecFeedde2026-08-31T08:02:45.561187+00:00Set-shifting Behavioral Test for Harnessed Agentshttps://blog.schoergendorfer.com/post/memorycard-topic-aware-multi-modal-clue-compression-for-long-video-question-answSecFeedde2026-08-31T08:02:45.485552+00:00MemoryCard: Topic-Aware Multi-Modal Clue Compression for Long-Video Question Answeringhttps://blog.schoergendorfer.com/post/a-wolf-in-sheeps-clothing-targeted-routing-hijacking-in-federated-ragSecFeedde2026-08-31T08:02:45.458451+00:00A Wolf in Sheep's Clothing: Targeted Routing Hijacking in Federated RAGhttps://blog.schoergendorfer.com/post/the-telephone-game-evaluating-semantic-drift-in-unified-modelsSecFeedde2026-08-31T08:02:45.431503+00:00The Telephone Game: Evaluating Semantic Drift in Unified Modelshttps://blog.schoergendorfer.com/post/long-story-short-story-level-video-understanding-from-20k-short-filmsSecFeedde2026-08-31T08:02:45.403884+00:00Long Story Short: Story-level Video Understanding from 20K Short Filmshttps://blog.schoergendorfer.com/post/prism-self-pruning-intrinsic-selection-method-for-training-free-multimodal-dataSecFeedde2026-08-31T08:02:45.376106+00:00PRISM: Self-Pruning Intrinsic Selection Method for Training-Free Multimodal Data Selectionhttps://blog.schoergendorfer.com/post/representing-and-parsing-korean-constituency-structure-at-different-levels-of-gr-2SecFeedde2026-08-31T08:02:45.349449+00:00Representing and Parsing Korean Constituency Structure at Different Levels of Granularityhttps://blog.schoergendorfer.com/post/surgical-alignment-in-knowledge-graph-training-for-clinical-diagnosis-with-large-2SecFeedde2026-08-31T08:02:45.323019+00:00Surgical Alignment in Knowledge Graph Training for Clinical Diagnosis with Large Language Modelshttps://blog.schoergendorfer.com/post/comparing-chunking-and-embedding-strategies-for-turkish-rag-systems-2SecFeedde2026-08-31T08:02:45.295742+00:00Comparing Chunking and Embedding Strategies for Turkish RAG Systemshttps://blog.schoergendorfer.com/post/self-generated-text-recognition-quality-heuristics-cross-task-transfer-and-downs-2SecFeedde2026-08-31T08:02:45.269641+00:00Self-Generated Text Recognition: Quality Heuristics, Cross-Task Transfer, and Downstream Bias in LLM Evaluationhttps://blog.schoergendorfer.com/post/ai-models-can-predict-and-collaboratively-modulate-human-memory-searchSecFeedde2026-08-31T08:02:45.243819+00:00AI Models Can Predict and Collaboratively Modulate Human Memory Searchhttps://blog.schoergendorfer.com/post/elementcheck-complexity-aware-long-form-text-factuality-evaluation-via-sentence-2SecFeedde2026-08-31T08:02:45.217291+00:00ElementCheck: Complexity-Aware Long-Form Text Factuality Evaluation via Sentence Elementshttps://blog.schoergendorfer.com/post/treegraft-adaptive-multi-drafter-grafting-for-tree-based-speculative-decoding-2SecFeedde2026-08-31T07:47:32.239172+00:00TreeGraft: Adaptive Multi-Drafter Grafting for Tree-Based Speculative Decodinghttps://blog.schoergendorfer.com/post/overview-of-shroom-visions-2026-a-shared-task-on-hallucination-detection-in-larg-2SecFeedde2026-08-31T07:47:32.212430+00:00Overview of SHROOM-Visions 2026: A Shared Task on Hallucination Detection in Large Vision-Language Modelshttps://blog.schoergendorfer.com/post/one-form-to-transfer-them-all-pretraining-multilingual-language-models-beyond-na-2SecFeedde2026-08-31T07:47:32.186327+00:00One Form to Transfer Them All: Pretraining Multilingual Language Models Beyond Native Orthographyhttps://blog.schoergendorfer.com/post/mathadv-what-theorem-provers-know-reason-formalize-and-generalize-2SecFeedde2026-08-31T07:47:32.160367+00:00MathAdv: What Theorem Provers Know, Reason, Formalize, and Generalizehttps://blog.schoergendorfer.com/post/a-declarative-procedural-perspective-on-expert-routing-in-bilingual-mixture-of-e-2SecFeedde2026-08-31T07:47:32.136458+00:00A Declarative-Procedural Perspective on Expert Routing in Bilingual Mixture-of-Experts Language Modelshttps://blog.schoergendorfer.com/post/why-knowing-both-hops-is-not-enough-understanding-two-hop-generalization-in-lang-3SecFeedde2026-08-31T07:47:32.089020+00:00Why Knowing Both Hops Is Not Enough: Understanding Two-Hop Generalization in Language Modelshttps://blog.schoergendorfer.com/post/profilefoundry-a-synthetic-person-object-substrate-for-privacy-memory-and-tool-uSecFeedde2026-08-31T07:47:32.065335+00:00ProfileFoundry: A Synthetic Person-Object Substrate for Privacy, Memory, and Tool-Use Evaluation in LLM Agenthttps://blog.schoergendorfer.com/post/persuasion-index-a-theory-guided-framework-for-persuasion-analysisSecFeedde2026-08-31T07:47:32.023937+00:00Persuasion Index: A Theory-Guided Framework for Persuasion Analysishttps://blog.schoergendorfer.com/post/does-finetuning-with-scientific-data-increase-hallucinations-a-multi-domain-factSecFeedde2026-08-31T07:47:32.000242+00:00Does Finetuning with Scientific Data Increase Hallucinations? A Multi-domain Factuality Evaluation of LLMshttps://blog.schoergendorfer.com/post/the-granularity-gap-a-multi-dimensional-cross-generational-audit-of-sycophancy-i-2SecFeedde2026-08-31T07:47:31.976412+00:00The Granularity Gap: A Multi-Dimensional Cross-Generational Audit of Sycophancy in Gemini Modelshttps://blog.schoergendorfer.com/post/self-evaluation-is-already-there-eliciting-latent-judge-calibration-in-base-llmsSecFeedde2026-08-31T07:47:31.953608+00:00Self-Evaluation Is Already There: Eliciting Latent Judge Calibration in Base LLMs with Minimal Datahttps://blog.schoergendorfer.com/post/realclawbench-live-openclaw-benchmarks-from-real-developer-agent-sessionsSecFeedde2026-08-31T07:47:31.933565+00:00RealClawBench: Live OpenClaw Benchmarks from Real Developer-Agent Sessionshttps://blog.schoergendorfer.com/post/cram-centroid-routing-and-adaptive-moe-for-multimodal-continual-instruction-tuni-2SecFeedde2026-08-31T07:47:31.913194+00:00CRAM: Centroid-Routing and Adaptive MoE for Multimodal Continual Instruction Tuninghttps://blog.schoergendorfer.com/post/benchmarking-llm-as-a-judge-for-long-form-output-evaluation-2SecFeedde2026-08-31T07:47:31.888640+00:00Benchmarking LLM-as-a-Judge for Long-Form Output Evaluationhttps://blog.schoergendorfer.com/post/diffusent-towards-a-unified-diffusion-framework-for-aspect-based-sentiment-analySecFeedde2026-08-31T07:47:31.859473+00:00DiffuSent: Towards a Unified Diffusion Framework for Aspect-Based Sentiment Analysishttps://blog.schoergendorfer.com/post/human-label-variation-as-stable-signal-learning-annotator-specific-explanation-bSecFeedde2026-08-31T07:47:31.815704+00:00Human Label Variation as Stable Signal: Learning Annotator-Specific Explanation Behavior via Cross-Annotator Preference Optimizationhttps://blog.schoergendorfer.com/post/why-are-all-llms-obsessed-with-japanese-culture-on-the-hidden-cultural-and-regioSecFeedde2026-08-31T07:32:32.754248+00:00Why are all LLMs Obsessed with Japanese Culture? On the Hidden Cultural and Regional Biases of LLMshttps://blog.schoergendorfer.com/post/select-label-evaluate-active-testing-in-nlpSecFeedde2026-08-31T07:32:32.729007+00:00Select, Label, Evaluate: Active Testing in NLPhttps://blog.schoergendorfer.com/post/large-reasoning-models-struggle-to-transfer-parametric-knowledge-across-scriptsSecFeedde2026-08-31T07:32:32.675506+00:00Large Reasoning Models Struggle to Transfer Parametric Knowledge Across Scriptshttps://blog.schoergendorfer.com/post/from-leaky-thoughts-to-private-reasoning-controlling-what-lrms-say-to-themselves-2SecFeedde2026-08-31T07:32:32.628145+00:00From Leaky Thoughts to Private Reasoning: Controlling What LRMs Say to Themselveshttps://blog.schoergendorfer.com/post/fence-a-financial-and-multimodal-jailbreak-detection-datasetSecFeedde2026-08-31T07:32:32.602259+00:00FENCE: A Financial and Multimodal Jailbreak Detection Datasethttps://blog.schoergendorfer.com/post/cofrgenet-continued-fraction-architectures-for-language-generationSecFeedde2026-08-31T07:32:32.577519+00:00CoFrGeNet: Continued Fraction Architectures for Language Generationhttps://blog.schoergendorfer.com/post/beyond-the-rabbit-hole-mapping-the-relational-harms-of-qanon-radicalizationSecFeedde2026-08-31T07:32:32.555023+00:00Beyond the Rabbit Hole: Mapping the Relational Harms of QAnon Radicalizationhttps://blog.schoergendorfer.com/post/omnifusion-simultaneous-multilingual-multimodal-translations-via-modular-fusionSecFeedde2026-08-31T07:32:32.509411+00:00OmniFusion: Simultaneous Multilingual Multimodal Translations via Modular Fusionhttps://blog.schoergendorfer.com/post/learning-a-single-token-to-replace-long-system-prompts-in-llmsSecFeedde2026-08-31T07:32:32.486516+00:00Learning a Single Token to Replace Long System Prompts in LLMshttps://blog.schoergendorfer.com/post/smrc-aligning-large-language-models-with-student-reasoning-for-mathematical-erroSecFeedde2026-08-31T07:32:32.441119+00:00SMRC: Aligning Large Language Models with Student Reasoning for Mathematical Error Correctionhttps://blog.schoergendorfer.com/post/quantifying-affective-bias-in-low-resource-media-large-scale-emotion-profiling-oSecFeedde2026-08-31T07:32:32.415937+00:00Quantifying Affective Bias in Low-Resource Media: Large-Scale Emotion Profiling of Bengali Headlineshttps://blog.schoergendorfer.com/post/prism-agentic-retrieval-with-llms-for-multi-hop-question-answeringSecFeedde2026-08-31T07:32:32.389402+00:00PRISM: Agentic Retrieval with LLMs for Multi-Hop Question Answeringhttps://blog.schoergendorfer.com/post/steering-multimodal-large-language-models-decoding-for-context-aware-safetySecFeedde2026-08-31T07:32:32.364550+00:00Steering Multimodal Large Language Models Decoding for Context-Aware Safetyhttps://blog.schoergendorfer.com/post/beyond-the-rosetta-stone-unification-forces-in-generalization-dynamics-2SecFeedde2026-08-31T07:32:32.337863+00:00Beyond the Rosetta Stone: Unification Forces in Generalization Dynamicshttps://blog.schoergendorfer.com/post/cognitive-chain-of-thought-cocot-structured-multimodal-reasoning-about-social-siSecFeedde2026-08-31T07:32:32.309042+00:00Cognitive Chain-of-Thought (CoCoT): Structured Multimodal Reasoning about Social Situationshttps://blog.schoergendorfer.com/post/pruning-laws-for-large-language-modelsSecFeedde2026-08-31T07:17:41.606407+00:00Pruning Laws for Large Language Modelshttps://blog.schoergendorfer.com/post/evaluating-the-performance-of-large-language-models-on-gaokao-benchmarkSecFeedde2026-08-31T07:17:41.582650+00:00Evaluating the Performance of Large Language Models on GAOKAO Benchmarkhttps://blog.schoergendorfer.com/post/beacon-behavior-anchored-cross-source-knowledge-graph-construction-for-cyber-thrSecFeedde2026-08-31T07:17:41.537396+00:00BEACON: Behavior-Anchored Cross-Source Knowledge Graph Construction for Cyber Threat Intelligencehttps://blog.schoergendorfer.com/post/camodocs-a-poisoning-attack-against-retrieval-augmented-language-models-using-caSecFeedde2026-08-31T07:17:41.511316+00:00CamoDocs: A Poisoning Attack Against Retrieval-Augmented Language Models Using Camouflaged Documentshttps://blog.schoergendorfer.com/post/semantic-head-specialization-guides-hybrid-vit-attention-for-multimodal-llmsSecFeedde2026-08-31T07:17:41.484862+00:00Semantic Head Specialization Guides Hybrid ViT Attention for Multimodal LLMshttps://blog.schoergendorfer.com/post/layered-llm-defenses-as-an-ensemble-access-tiers-inference-cost-and-the-measuredSecFeedde2026-08-31T07:17:41.457708+00:00Layered LLM Defenses as an Ensemble: Access Tiers, Inference Cost, and the Measured Failure Correlation Between Defense Layershttps://blog.schoergendorfer.com/post/aim-anchor-identity-features-then-match-for-multimodal-large-language-model-unleSecFeedde2026-08-31T07:17:41.425677+00:00AIM: Anchor Identity Features, Then Match for Multimodal Large Language Model Unlearninghttps://blog.schoergendorfer.com/post/synth-jdoc-synthesizing-a-japanese-document-image-dataset-for-ocr-with-diverse-lSecFeedde2026-08-31T07:17:41.396531+00:00Synth-JDoc: Synthesizing a Japanese Document Image Dataset for OCR with Diverse Layouts and Embedded Imageshttps://blog.schoergendorfer.com/post/benchmarking-large-language-model-agent-societies-against-human-behavioural-distSecFeedde2026-08-31T07:17:41.368293+00:00Benchmarking large language model agent societies against human behavioural distributionshttps://blog.schoergendorfer.com/post/speculative-probing-llm-monitoring-at-speculative-decoding-costSecFeedde2026-08-31T07:17:41.336530+00:00Speculative Probing: LLM Monitoring at Speculative-Decoding Costhttps://blog.schoergendorfer.com/post/discti-who-needs-to-know-timely-automated-sector-aware-cyber-threat-intelligenceSecFeedde2026-08-31T07:17:41.309761+00:00DisCTI: Who Needs to Know Timely? Automated Sector-Aware Cyber Threat Intelligence Disseminationhttps://blog.schoergendorfer.com/post/auditing-generative-audio-calls-for-known-task-audio-llm-evaluationSecFeedde2026-08-31T07:17:41.241754+00:00Auditing Generative Audio Calls for Known-Task Audio-LLM Evaluationhttps://blog.schoergendorfer.com/post/cedar-automata-as-verifiable-interfaces-for-language-guided-embodied-actionSecFeedde2026-08-31T07:17:41.213711+00:00CEDAR: Automata as Verifiable Interfaces for Language-Guided Embodied Actionhttps://blog.schoergendorfer.com/post/why-didnt-it-check-unsupported-final-claims-and-their-repair-in-two-tool-equippeSecFeedde2026-08-31T07:17:41.191653+00:00Why Didn't It Check? Unsupported Final Claims and Their Repair in Two Tool-Equipped Language Modelshttps://blog.schoergendorfer.com/post/semantic-watermarking-with-order-robust-detection-over-sub-sentence-unitsSecFeedde2026-08-31T07:17:41.145346+00:00Semantic Watermarking with Order-Robust Detection over Sub-sentence Unitshttps://blog.schoergendorfer.com/post/retrieving-relations-detecting-fallacies-a-rag-approach-to-political-debate-analSecFeedde2026-08-31T07:17:41.121629+00:00Retrieving Relations, Detecting Fallacies: A RAG Approach to Political Debate Analysishttps://blog.schoergendorfer.com/post/a-formal-limitation-on-learning-human-language-from-textual-corporaSecFeedde2026-08-31T07:03:10.589545+00:00A Formal Limitation on Learning Human Language From Textual Corporahttps://blog.schoergendorfer.com/post/ladders-in-chaos-when-how-and-perhaps-why-does-test-time-scaling-improve-llm-macSecFeedde2026-08-31T07:03:10.566040+00:00Ladders in Chaos: When, How, (and Perhaps Why) Does Test-Time Scaling Improve LLM Machine Translationhttps://blog.schoergendorfer.com/post/phoneme-and-word-level-metrics-using-self-supervised-speech-representations-forSecFeedde2026-08-31T07:03:10.543786+00:00Phoneme- and Word-Level Metrics Using Self-Supervised Speech Representations for Forced Alignment Evaluationhttps://blog.schoergendorfer.com/post/nl2agbench-benchmarking-llm-auto-formalization-for-alphageometrySecFeedde2026-08-31T07:03:10.520637+00:00NL2AGBench: Benchmarking LLM Auto-Formalization for AlphaGeometryhttps://blog.schoergendorfer.com/post/blind-men-and-the-elephant-probing-the-epistemic-myopia-of-llms-under-long-tailSecFeedde2026-08-31T07:03:10.497705+00:00Blind Men and the Elephant: Probing the Epistemic Myopia of LLMs under Long-Tail Divergent Knowledgehttps://blog.schoergendorfer.com/post/contextpilot-teaching-agents-for-proactive-context-management-via-fine-grained-rSecFeedde2026-08-31T07:03:10.475098+00:00ContextPilot: Teaching Agents for Proactive Context Management via Fine-grained RLhttps://blog.schoergendorfer.com/post/fidelity-is-not-enough-dispatch-level-instrumentation-for-agentic-datasheet-extrSecFeedde2026-08-31T07:03:10.451328+00:00Fidelity Is Not Enough: Dispatch-Level Instrumentation for Agentic Datasheet Extractionhttps://blog.schoergendorfer.com/post/stranger-fan-or-peer-a-systematic-study-on-the-role-of-interlocutor-in-persona-bSecFeedde2026-08-31T07:03:10.428686+00:00Stranger, Fan, or Peer? A Systematic Study on the Role of Interlocutor in Persona-Based Dialogue Generationhttps://blog.schoergendorfer.com/post/are-these-modules-worth-their-cost-a-paradigm-level-accuracy-cost-analysis-of-inSecFeedde2026-08-31T07:03:10.404577+00:00Are These Modules Worth Their Cost? A Paradigm-Level Accuracy-Cost Analysis of In-context Learning Text-to-SQLhttps://blog.schoergendorfer.com/post/a-unified-framework-to-elicit-structured-feedback-for-interpretable-multi-traitSecFeedde2026-08-31T07:03:10.381355+00:00A Unified Framework to Elicit Structured Feedback for Interpretable Multi-Trait Essay Scoringhttps://blog.schoergendorfer.com/post/cultureconverse-a-multilingual-multi-turn-simulation-harness-for-culturally-grouSecFeedde2026-08-31T07:03:10.353845+00:00CultureConverse: A Multilingual Multi-turn Simulation Harness for Culturally Grounded Assistance in East and Southeast Asiahttps://blog.schoergendorfer.com/post/when-linguistic-and-internal-confidence-diverge-in-large-language-modelsSecFeedde2026-08-31T07:03:10.329540+00:00When Linguistic and Internal Confidence Diverge in Large Language Modelshttps://blog.schoergendorfer.com/post/personaforge-realistic-multi-turn-user-simulation-for-agentic-systemsSecFeedde2026-08-31T07:03:10.305214+00:00PersonaForge: Realistic Multi-Turn User Simulation for Agentic Systemshttps://blog.schoergendorfer.com/post/a-probabilistic-interpretation-of-kv-cache-evictionSecFeedde2026-08-31T07:03:10.281909+00:00A Probabilistic Interpretation of KV Cache Evictionhttps://blog.schoergendorfer.com/post/embedding-models-for-stance-aware-argument-retrievalSecFeedde2026-08-31T07:03:10.259527+00:00Embedding Models for Stance-Aware Argument Retrievalhttps://blog.schoergendorfer.com/post/text-restoration-of-ancient-documents-with-language-modelsSecFeedde2026-08-31T07:03:10.235936+00:00Text Restoration of Ancient Documents with Language Modelshttps://blog.schoergendorfer.com/post/nested-byte-level-vocabularies-are-cheap-to-deploy-and-expensive-to-share-a-preSecFeedde2026-08-31T07:03:10.208648+00:00Nested Byte-Level Vocabularies Are Cheap to Deploy and Expensive to Share: A Pre-Registered Negative Resulthttps://blog.schoergendorfer.com/post/finexam-10k-when-retrieval-helps-financial-reasoningSecFeedde2026-08-31T07:03:10.185848+00:00FinExam-10K: When Retrieval Helps Financial Reasoning?https://blog.schoergendorfer.com/post/h-scale-hessian-guided-scale-refinement-for-nvfp4-sub-byte-llm-inferenceSecFeedde2026-08-31T07:03:10.164531+00:00H-Scale: Hessian-Guided Scale Refinement for NVFP4 Sub-Byte LLM Inferencehttps://blog.schoergendorfer.com/post/cneo-bench-diagnosing-large-language-models-on-chinese-neologismsSecFeedde2026-08-31T07:03:10.141588+00:00CNeo-Bench: Diagnosing Large Language Models on Chinese Neologismshttps://blog.schoergendorfer.com/post/simpcue-cue-based-prompting-for-multilingual-text-simplificationSecFeedde2026-08-31T06:47:37.946188+00:00SimpCue: Cue-Based Prompting for Multilingual Text Simplificationhttps://blog.schoergendorfer.com/post/beyond-global-scalars-synergizing-token-level-statistics-and-deep-semantics-forSecFeedde2026-08-31T06:47:37.894827+00:00Beyond Global Scalars: Synergizing Token-Level Statistics and Deep Semantics for Adversarial AIGC Text Detectionhttps://blog.schoergendorfer.com/post/predicting-turn-taking-outcomes-in-multi-party-conversation-interpretable-modellSecFeedde2026-08-31T06:47:37.869663+00:00Predicting Turn-Taking Outcomes in Multi-Party Conversation: Interpretable Modelling of Speech and Gaze Dynamics with Interpersonal Closenesshttps://blog.schoergendorfer.com/post/quorum-quality-optimized-routing-using-multiple-annotatorsSecFeedde2026-08-31T06:47:37.845670+00:00QUORUM: QUality-Optimized Routing Using Multiple annotatorshttps://blog.schoergendorfer.com/post/entity-memory-graph-retrieval-improves-evidence-coverage-in-long-conversation-quSecFeedde2026-08-31T06:47:37.792818+00:00Entity-Memory Graph Retrieval Improves Evidence Coverage in Long-Conversation Question Answeringhttps://blog.schoergendorfer.com/post/what-makes-agent-memory-useful-for-reliable-unanswerable-question-handlingSecFeedde2026-08-31T06:47:37.761766+00:00What Makes Agent Memory Useful for Reliable Unanswerable Question Handling?https://blog.schoergendorfer.com/post/landingagent-a-reference-annotated-dataset-and-agentic-generation-framework-forSecFeedde2026-08-31T06:47:37.732118+00:00LandingAgent: A Reference-Annotated Dataset and Agentic Generation Framework for Landing Pageshttps://blog.schoergendorfer.com/post/ai-writers-have-a-consistent-stylometric-footprint-but-ai-editors-do-notSecFeedde2026-08-31T06:47:37.707943+00:00AI Writers Have a Consistent Stylometric Footprint, but AI Editors Do Nothttps://blog.schoergendorfer.com/post/evoharmbench-breaking-content-moderation-with-iterative-human-like-evasionSecFeedde2026-08-31T06:47:37.679420+00:00EvoHarmBench: Breaking Content Moderation with Iterative Human-Like Evasionhttps://blog.schoergendorfer.com/post/synthetic-linguistic-agency-how-an-embodied-mortal-agent-learns-linguistic-afforSecFeedde2026-08-31T06:47:37.649644+00:00Synthetic Linguistic Agency: How an Embodied Mortal Agent Learns Linguistic Affordances through Consequential Social Experiencehttps://blog.schoergendorfer.com/post/personaedit-representative-sample-selection-for-personalized-model-editingSecFeedde2026-08-31T06:47:37.622905+00:00PersonaEdit: Representative Sample Selection for Personalized Model Editinghttps://blog.schoergendorfer.com/post/representation-of-syntax-in-llms-through-the-lens-of-linear-distance-and-similarSecFeedde2026-08-31T06:47:37.598694+00:00Representation of syntax in LLMs through the lens of linear distance and similarity-aware entropyhttps://blog.schoergendorfer.com/post/compositional-failure-in-audio-visual-llms-late-layer-prior-dominance-under-crosSecFeedde2026-08-31T06:47:37.573154+00:00Compositional Failure in Audio-Visual LLMs: Late-Layer Prior Dominance Under Cross-modal Conflicthttps://blog.schoergendorfer.com/post/load-bearing-context-the-question-damage-score-for-evaluating-context-reliance-iSecFeedde2026-08-31T06:47:37.521665+00:00Load-Bearing Context: The Question Damage Score for Evaluating Context Reliance in Linguistic Reasoninghttps://blog.schoergendorfer.com/post/below-the-noise-floor-bimodal-seed-collapse-and-distinct-failure-modes-in-smallSecFeedde2026-08-31T06:47:37.493132+00:00Below the Noise Floor: Bimodal Seed Collapse and Distinct Failure Modes in Small-Model Knowledge Distillationhttps://blog.schoergendorfer.com/post/first-make-it-playable-then-make-it-good-staged-interaction-learning-for-small-dSecFeedde2026-08-31T06:47:37.467408+00:00First Make It Playable, Then Make It Good: Staged Interaction Learning for Small Dialogue-Game Agentshttps://blog.schoergendorfer.com/post/knowing-before-answering-decoding-language-models-for-reliable-ragSecFeedde2026-08-31T06:47:37.438023+00:00Knowing Before Answering: Decoding Language Models for Reliable RAGhttps://blog.schoergendorfer.com/post/when-tokenizers-fail-byte-level-chunking-for-zero-shot-transfer-to-low-resourceSecFeedde2026-08-31T06:32:23.483241+00:00When Tokenizers Fail: Byte-Level Chunking for Zero-Shot Transfer to Low-Resource Languageshttps://blog.schoergendorfer.com/post/trajectory-level-speculative-decoding-for-diffusion-language-modelsSecFeedde2026-08-31T06:32:23.460665+00:00Trajectory-Level Speculative Decoding for Diffusion Language Modelshttps://blog.schoergendorfer.com/post/a-survey-on-rubric-guided-reinforcement-learning-for-language-modelsSecFeedde2026-08-31T06:32:23.438269+00:00A Survey on Rubric-Guided Reinforcement Learning for Language Modelshttps://blog.schoergendorfer.com/post/inspire-an-internalize-then-improve-approach-for-example-driven-mathematical-reaSecFeedde2026-08-31T06:32:23.414900+00:00INSPIRE: An Internalize-Then-Improve Approach for Example-Driven Mathematical Reasoninghttps://blog.schoergendorfer.com/post/xhotpotqa-a-benchmark-for-cross-lingual-knowledge-composition-in-multi-hop-questSecFeedde2026-08-31T06:32:23.390759+00:00XHotpotQA: A Benchmark for Cross-Lingual Knowledge Composition in Multi-Hop Question Answeringhttps://blog.schoergendorfer.com/post/select-dont-train-the-benefits-of-modular-entity-disambiguation-with-llm-based-sSecFeedde2026-08-31T06:32:23.368753+00:00Select, Don't Train: The Benefits of Modular Entity Disambiguation with LLM-Based Selectionhttps://blog.schoergendorfer.com/post/uic-aihealth4all-at-archehr-qa-2026-answer-first-evidence-grounding-for-clinicalSecFeedde2026-08-31T06:32:23.344714+00:00UIC-AIHealth4All at ArchEHR-QA 2026: Answer-First Evidence Grounding for Clinical Question Answeringhttps://blog.schoergendorfer.com/post/pace-publisher-adaptive-content-extraction-via-agentic-automationSecFeedde2026-08-31T06:32:23.321533+00:00PACE: Publisher-Adaptive Content Extraction via Agentic Automationhttps://blog.schoergendorfer.com/post/the-effect-of-emotional-context-on-large-language-models-endorsement-of-prematurSecFeedde2026-08-31T06:32:23.298268+00:00The Effect of Emotional Context on Large Language Models' Endorsement of Premature Decisions: Comparing Emotional Vulnerability Across Six Commercial Modelshttps://blog.schoergendorfer.com/post/sledgehammer-or-scalpel-a-fine-grained-adaptive-framework-for-implicit-hate-speeSecFeedde2026-08-31T06:32:23.275307+00:00Sledgehammer or Scalpel? A Fine-grained Adaptive Framework for Implicit Hate Speechhttps://blog.schoergendorfer.com/post/semantic-overlays-mitigating-prompt-injection-with-annotations-beyond-tokens-and-2SecFeedde2026-08-31T06:32:23.253263+00:00Semantic Overlays: Mitigating Prompt Injection with Annotations Beyond Tokens and Steering Vectorshttps://blog.schoergendorfer.com/post/gan-diff-coupling-pretrained-wgan-gp-features-with-conditional-diffusion-u-nets-2SecFeedde2026-08-31T06:32:23.228696+00:00GAN-Diff : Coupling Pretrained WGAN-GP Features with Conditional Diffusion U-Netshttps://blog.schoergendorfer.com/post/what-neural-network-field-theory-can-and-cannot-realise-on-a-computerSecFeedde2026-08-31T06:32:23.204678+00:00What Neural Network Field Theory Can and Cannot Realise on a Computerhttps://blog.schoergendorfer.com/post/juryprobe-an-empirical-consensus-risk-diagnostic-for-routing-reference-free-fact-2SecFeedde2026-08-31T06:32:23.181552+00:00JuryProbe: An Empirical Consensus-Risk Diagnostic for Routing Reference-Free Factuality Judge Panels to Grounded Verificationhttps://blog.schoergendorfer.com/post/establishing-boundary-kkt-convergence-of-mirror-descent-through-reparameterizatiSecFeedde2026-08-31T06:32:23.155464+00:00Establishing Boundary KKT Convergence of Mirror Descent through Reparameterizationhttps://blog.schoergendorfer.com/post/where-steering-signals-come-from-activation-source-selection-in-activation-steer-2SecFeedde2026-08-31T06:32:23.133665+00:00Where Steering Signals Come From: Activation Source Selection in Activation Steeringhttps://blog.schoergendorfer.com/post/robust-chance-constrained-optimization-using-a-continuous-parameter-space-wasser-2SecFeedde2026-08-31T06:32:23.109677+00:00Robust Chance-Constrained Optimization using a Continuous Parameter Space Wasserstein-2 Ambiguity Set of Gaussian Mixtureshttps://blog.schoergendorfer.com/post/an-end-to-end-hybrid-quantum-classical-sampling-workflow-for-discrete-markov-ran-2SecFeedde2026-08-31T06:32:23.083312+00:00An End-to-End Hybrid Quantum--Classical Sampling Workflow for Discrete Markov Random Fields: A Reproducible Case Studyhttps://blog.schoergendorfer.com/post/closing-the-operational-gap-in-semantic-cachingSecFeedde2026-08-31T06:32:23.059307+00:00Closing the Operational Gap in Semantic Cachinghttps://blog.schoergendorfer.com/post/tokenpilot-cache-efficient-context-management-for-llm-agentsSecFeedde2026-08-31T06:32:23.031690+00:00TokenPilot: Cache-Efficient Context Management for LLM Agentshttps://blog.schoergendorfer.com/post/recoursebench-a-modular-framework-for-reproducible-algorithmic-recourse-evaluatiSecFeedde2026-08-31T06:17:27.409817+00:00RecourseBench: A Modular Framework for Reproducible Algorithmic Recourse Evaluationhttps://blog.schoergendorfer.com/post/toolsense-a-diagnostic-framework-for-auditing-parametric-tool-knowledge-in-llmsSecFeedde2026-08-31T06:17:27.385776+00:00ToolSense: A Diagnostic Framework for Auditing Parametric Tool Knowledge in LLMshttps://blog.schoergendorfer.com/post/wino-a-weak-form-physics-informed-neural-operator-for-hyperelasticity-on-variabl-2SecFeedde2026-08-31T06:17:27.360729+00:00WINO: A Weak-Form Physics Informed Neural Operator for Hyperelasticity on Variable Domainshttps://blog.schoergendorfer.com/post/skillsafetybench-evaluating-agent-safety-under-skill-facing-attack-surfacesSecFeedde2026-08-31T06:17:27.309983+00:00SkillSafetyBench: Evaluating Agent Safety under Skill-Facing Attack Surfaceshttps://blog.schoergendorfer.com/post/d3-gym-constructing-real-world-verifiable-environments-for-data-driven-discoverySecFeedde2026-08-31T06:17:27.283919+00:00D3-Gym: Constructing Real-World Verifiable Environments for Data-Driven Discoveryhttps://blog.schoergendorfer.com/post/prompts-without-evidence-how-neuroimaging-mentions-shift-clinical-vision-languagSecFeedde2026-08-31T06:17:27.196406+00:00Prompts Without Evidence: How Neuroimaging Mentions Shift Clinical Vision-Language Model Predictionshttps://blog.schoergendorfer.com/post/deflation-pinns-learning-multiple-solutions-for-pdes-and-landau-de-gennesSecFeedde2026-08-31T06:17:27.171675+00:00Deflation-PINNs: Learning Multiple Solutions for PDEs and Landau-de Genneshttps://blog.schoergendorfer.com/post/camera-agnostic-pruning-of-3d-gaussian-splats-via-descriptor-based-beta-evidence-2SecFeedde2026-08-31T06:17:27.147420+00:00Camera-Agnostic Pruning of 3D Gaussian Splats via Descriptor-Based Beta Evidencehttps://blog.schoergendorfer.com/post/the-autonomy-tax-defense-training-breaks-llm-agentsSecFeedde2026-08-31T06:17:27.123449+00:00The Autonomy Tax: Defense Training Breaks LLM Agentshttps://blog.schoergendorfer.com/post/agentic-kube-a-graph-enhanced-multi-agent-reinforcement-learning-framework-for-mSecFeedde2026-08-31T06:17:27.095631+00:00Agentic-Kube: A Graph-Enhanced Multi-Agent Reinforcement Learning Framework for Multi-Objective Kubernetes Schedulinghttps://blog.schoergendorfer.com/post/flowcorrect-efficient-interactive-correction-of-generative-flow-policies-for-robSecFeedde2026-08-31T06:17:27.067331+00:00FlowCorrect: Efficient Interactive Correction of Generative Flow Policies for Robotic Manipulationhttps://blog.schoergendorfer.com/post/mine-and-refine-optimizing-graded-relevance-in-e-commerce-semantic-search-retrieSecFeedde2026-08-31T06:17:27.041604+00:00Mine and Refine: Optimizing Graded Relevance in E-commerce Semantic Search Retrievalhttps://blog.schoergendorfer.com/post/robust-assortment-optimization-from-observational-dataSecFeedde2026-08-31T06:17:27.016189+00:00Robust Assortment Optimization from Observational Datahttps://blog.schoergendorfer.com/post/scale-self-uncertainty-conditioned-adaptive-looking-and-execution-for-vision-lanSecFeedde2026-08-31T06:17:26.991730+00:00SCALE: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Modelshttps://blog.schoergendorfer.com/post/learning-fast-monomial-orders-for-grobner-basis-computationsSecFeedde2026-08-31T06:17:26.970874+00:00Learning Fast Monomial Orders for Gr\"obner Basis Computationshttps://blog.schoergendorfer.com/post/aligning-agentic-world-models-via-knowledgeable-experience-learningSecFeedde2026-08-31T06:17:26.949213+00:00Aligning Agentic World Models via Knowledgeable Experience Learninghttps://blog.schoergendorfer.com/post/prequential-posteriorsSecFeedde2026-08-31T06:03:56.039578+00:00Prequential posteriorshttps://blog.schoergendorfer.com/post/think-at-hard-dynamic-looped-transformers-for-improved-reasoningSecFeedde2026-08-31T06:03:56.013386+00:00Think-at-Hard: Dynamic Looped Transformers for Improved Reasoninghttps://blog.schoergendorfer.com/post/multilingual-lexical-feature-analysis-of-spoken-language-for-predicting-major-deSecFeedde2026-08-31T06:03:55.992425+00:00Multilingual Lexical Feature Analysis of Spoken Language for Predicting Major Depression Symptom Severityhttps://blog.schoergendorfer.com/post/examining-the-robustness-of-physics-informed-neural-networks-to-noise-for-inversSecFeedde2026-08-31T06:03:55.950720+00:00Examining the robustness of Physics-Informed Neural Networks to noise for Inverse Problemshttps://blog.schoergendorfer.com/post/oceangym-a-benchmark-environment-for-underwater-embodied-agentsSecFeedde2026-08-31T06:03:55.927765+00:00OceanGym: A Benchmark Environment for Underwater Embodied Agentshttps://blog.schoergendorfer.com/post/probabilistic-symbolic-regression-for-equation-discovery-via-operator-induced-anSecFeedde2026-08-31T06:03:55.902924+00:00Probabilistic Symbolic Regression for Equation Discovery via Operator-induced and Regularized Symbolic Forestshttps://blog.schoergendorfer.com/post/automatic-pronunciation-error-detection-and-correction-of-the-holy-qurans-learneSecFeedde2026-08-31T06:03:55.876647+00:00Automatic Pronunciation Error Detection and Correction of the Holy Quran's Learners Using Deep Learninghttps://blog.schoergendorfer.com/post/ampere-communication-efficient-and-high-accuracy-split-federated-learningSecFeedde2026-08-31T06:03:55.850545+00:00Ampere: Communication-Efficient and High-Accuracy Split Federated Learninghttps://blog.schoergendorfer.com/post/mixture-of-multicenter-experts-in-multimodal-ai-for-debiased-radiotherapy-targetSecFeedde2026-08-31T06:03:55.800446+00:00Mixture of Multicenter Experts in Multimodal AI for Debiased Radiotherapy Target Delineationhttps://blog.schoergendorfer.com/post/understanding-evolution-strategies-for-llm-reasoning-broader-reasoning-coverage-2SecFeedde2026-08-31T06:03:55.750097+00:00Understanding Evolution Strategies for LLM Reasoning: Broader Reasoning Coverage than GRPOhttps://blog.schoergendorfer.com/post/accurate-prediction-is-not-profitable-advice-profit-based-evaluation-of-machineSecFeedde2026-08-31T06:03:55.722572+00:00Accurate prediction is not profitable advice: profit-based evaluation of machine learning nitrogen recommendations in winter wheathttps://blog.schoergendorfer.com/post/neural-regression-with-embeddings-for-numerical-attribute-prediction-in-knowledg-2SecFeedde2026-08-31T06:03:55.693835+00:00Neural Regression with Embeddings for Numerical Attribute Prediction in Knowledge Graphshttps://blog.schoergendorfer.com/post/trust-the-mass-forced-weights-in-kv-cache-eviction-2SecFeedde2026-08-31T06:03:55.666986+00:00Trust the Mass: Forced Weights in KV-Cache Evictionhttps://blog.schoergendorfer.com/post/traceml-an-empirical-analysis-of-human-agent-planning-in-machine-learning-develo-2SecFeedde2026-08-31T06:03:55.633238+00:00TraceML: An Empirical Analysis of Human-Agent Planning in Machine Learning Developmenthttps://blog.schoergendorfer.com/post/on-policy-distillation-with-verifiable-reward-2SecFeedde2026-08-31T06:03:55.608319+00:00On-policy Distillation with Verifiable Rewardhttps://blog.schoergendorfer.com/post/jepa-x-cross-predictive-physics-grounding-for-forecastable-latent-dynamics-2SecFeedde2026-08-31T06:03:55.583229+00:00JEPA-x: Cross-Predictive Physics Grounding for Forecastable Latent Dynamicshttps://blog.schoergendorfer.com/post/ribospan-a-long-context-rna-foundation-model-for-versatile-rna-modelingSecFeedde2026-08-31T06:03:55.558200+00:00RIBOSPAN: A Long-Context RNA Foundation Model for Versatile RNA Modelinghttps://blog.schoergendorfer.com/post/how-architecture-and-training-affect-tpc-representations-across-experimentsSecFeedde2026-08-31T05:47:49.618907+00:00How Architecture and Training Affect TPC Representations Across Experimentshttps://blog.schoergendorfer.com/post/how-far-should-tokenization-go-predictive-effectiveness-and-relational-losslessnSecFeedde2026-08-31T05:47:49.595204+00:00How Far Should Tokenization Go? Predictive Effectiveness and Relational Losslessnesshttps://blog.schoergendorfer.com/post/recirculation-2SecFeedde2026-08-31T05:47:49.570021+00:00Recirculationhttps://blog.schoergendorfer.com/post/ed-csp-crystal-structure-prediction-from-electron-diffraction-3SecFeedde2026-08-31T05:47:49.542599+00:00ED-CSP: Crystal Structure Prediction from Electron Diffractionhttps://blog.schoergendorfer.com/post/locked-evaluation-surfaces-transfer-failure-and-sampling-depth-entanglement-in-cSecFeedde2026-08-31T05:47:49.516456+00:00Locked Evaluation Surfaces: Transfer Failure and Sampling-Depth Entanglement in CRISPRi Perturbation-Effect Predictionhttps://blog.schoergendorfer.com/post/specgradfilter-a-spectral-gradient-filtering-framework-for-taming-federated-hete-2SecFeedde2026-08-31T05:47:49.490521+00:00SpecGradFilter: A Spectral Gradient Filtering Framework for Taming Federated Heterogeneityhttps://blog.schoergendorfer.com/post/on-the-depth-scalability-of-logic-gate-networks-3SecFeedde2026-08-31T05:47:49.465053+00:00On the Depth Scalability of Logic Gate Networkshttps://blog.schoergendorfer.com/post/the-discrete-log-clock-how-a-transformer-learns-modular-multiplicationSecFeedde2026-08-31T05:47:49.433154+00:00The Discrete-Log Clock: How a Transformer Learns Modular Multiplicationhttps://blog.schoergendorfer.com/post/erp-xttn-interpretable-prototype-guided-cross-attention-for-cross-subject-erp-clSecFeedde2026-08-31T05:47:49.405068+00:00ERP-XTTN: Interpretable Prototype-Guided Cross-Attention for Cross-Subject ERP Classificationhttps://blog.schoergendorfer.com/post/longds-bench-on-the-failure-of-long-horizon-agentic-data-analysisSecFeedde2026-08-31T05:47:49.355475+00:00LongDS-Bench: On the Failure of Long-Horizon Agentic Data Analysishttps://blog.schoergendorfer.com/post/negligible-in-size-significant-in-effect-on-scale-vectors-in-large-language-modeSecFeedde2026-08-31T05:47:49.328406+00:00Negligible in Size, Significant in Effect: On Scale Vectors in Large Language Modelshttps://blog.schoergendorfer.com/post/more-expressive-feedforward-layers-part-i-token-adaptive-mixing-of-activationsSecFeedde2026-08-31T05:47:49.301307+00:00More Expressive Feedforward Layers: Part I. Token-Adaptive Mixing of Activationshttps://blog.schoergendorfer.com/post/learned-relay-representations-for-forward-thinking-discrete-diffusion-modelsSecFeedde2026-08-31T05:47:49.275099+00:00Learned Relay Representations for Forward-Thinking Discrete Diffusion Modelshttps://blog.schoergendorfer.com/post/implicitterrainv2-wavelet-guided-spatially-adaptive-neural-terrain-representatio-2SecFeedde2026-08-31T05:47:49.250803+00:00ImplicitTerrainV2: Wavelet-Guided Spatially Adaptive Neural Terrain Representationhttps://blog.schoergendorfer.com/post/abc-any-subset-autoregression-via-non-markovian-diffusion-bridges-in-continuousSecFeedde2026-08-31T05:47:49.223264+00:00ABC: Any-Subset Autoregression via Non-Markovian Diffusion Bridges in Continuous Time and Spacehttps://blog.schoergendorfer.com/post/semenrich-self-supervised-semantic-enrichment-of-radiology-reports-for-vision-laSecFeedde2026-08-31T05:47:49.198073+00:00SemEnrich: Self-Supervised Semantic Enrichment of Radiology Reports for Vision-Language Learninghttps://blog.schoergendorfer.com/post/budget-constrained-causal-bandits-bridging-uplift-modeling-and-sequential-decisiSecFeedde2026-08-31T05:47:49.171814+00:00Budget-Constrained Causal Bandits: Bridging Uplift Modeling and Sequential Decision-Makinghttps://blog.schoergendorfer.com/post/policylong-towards-on-policy-context-extensionSecFeedde2026-08-31T05:47:49.146341+00:00PolicyLong: Towards On-Policy Context Extensionhttps://blog.schoergendorfer.com/post/var-jepa-a-variational-formulation-of-the-joint-embedding-predictive-architecturSecFeedde2026-08-31T05:47:49.116418+00:00Var-JEPA: A Variational Formulation of the Joint-Embedding Predictive Architecture - Bridging Predictive and Generative Self-Supervised Learninghttps://blog.schoergendorfer.com/post/infomamba-an-attention-free-hybrid-mamba-transformer-modelSecFeedde2026-08-31T05:32:38.194082+00:00InfoMamba: An Attention-Free Hybrid Mamba-Transformer Modelhttps://blog.schoergendorfer.com/post/bayesian-experimental-design-for-model-discrepancy-calibration-a-rivalry-betweenSecFeedde2026-08-31T05:32:38.145650+00:00Bayesian Experimental Design for Model Discrepancy Calibration: A Rivalry between Kullback--Leibler Divergence and Wasserstein Distancehttps://blog.schoergendorfer.com/post/the-instability-of-safety-how-random-seeds-and-temperature-expose-inconsistent-lSecFeedde2026-08-31T05:32:38.119378+00:00The Instability of Safety: How Random Seeds and Temperature Expose Inconsistent LLM Refusal Behaviorhttps://blog.schoergendorfer.com/post/aspiration-based-perturbed-learning-automata-in-games-with-noisy-utility-measureSecFeedde2026-08-31T05:32:38.091691+00:00Aspiration-based Perturbed Learning Automata in Games with Noisy Utility Measurements. Part A: Stochastic Stability in Non-zero-Sum Gameshttps://blog.schoergendorfer.com/post/large-reasoning-models-learn-better-alignment-from-flawed-thinkingSecFeedde2026-08-31T05:32:38.060750+00:00Large Reasoning Models Learn Better Alignment from Flawed Thinkinghttps://blog.schoergendorfer.com/post/one-model-for-all-universal-pre-training-for-eeg-based-emotion-recognition-acrosSecFeedde2026-08-31T05:32:38.034798+00:00One Model for All: Universal Pre-training for EEG based Emotion Recognition across Heterogeneous Datasets and Paradigmshttps://blog.schoergendorfer.com/post/shift-before-you-learn-enabling-low-rank-representations-in-reinforcement-learniSecFeedde2026-08-31T05:32:38.009783+00:00Shift Before You Learn: Enabling Low-Rank Representations in Reinforcement Learninghttps://blog.schoergendorfer.com/post/class-incremental-continual-learning-with-self-organizing-maps-and-synthetic-repSecFeedde2026-08-31T05:32:37.984371+00:00Class Incremental Continual Learning with Self-Organizing Maps and Synthetic Replayhttps://blog.schoergendorfer.com/post/attention-as-conditioning-what-classical-learning-theory-predicts-about-linear-tSecFeedde2026-08-31T05:32:37.956028+00:00Attention as Conditioning: What Classical Learning Theory Predicts About Linear Transformershttps://blog.schoergendorfer.com/post/regcl-compact-continual-sam-adaptation-for-visual-grounding-in-multi-sensorial-mSecFeedde2026-08-31T05:32:37.928010+00:00RegCL: Compact Continual SAM Adaptation for Visual Grounding in Multi-Sensorial Mediahttps://blog.schoergendorfer.com/post/meta-prompt-optimization-for-llm-based-sequential-decision-makingSecFeedde2026-08-31T05:32:37.898034+00:00Meta-Prompt Optimization for LLM-Based Sequential Decision Makinghttps://blog.schoergendorfer.com/post/amortizing-intractable-inference-in-diffusion-models-for-vision-language-and-conSecFeedde2026-08-31T05:32:37.870318+00:00Amortizing intractable inference in diffusion models for vision, language, and controlhttps://blog.schoergendorfer.com/post/biases-in-expected-goals-models-confound-finishing-abilitySecFeedde2026-08-31T05:32:37.815402+00:00Biases in Expected Goals Models Confound Finishing Abilityhttps://blog.schoergendorfer.com/post/joint-bayesian-inference-of-graphical-structure-and-parameters-with-a-single-genSecFeedde2026-08-31T05:32:37.785558+00:00Joint Bayesian Inference of Graphical Structure and Parameters with a Single Generative Flow Networkhttps://blog.schoergendorfer.com/post/let-the-flows-tell-solving-graph-combinatorial-optimization-problems-with-gflownSecFeedde2026-08-31T05:32:37.756034+00:00Let the Flows Tell: Solving Graph Combinatorial Optimization Problems with GFlowNetshttps://blog.schoergendorfer.com/post/transformer-based-autonomous-driving-models-and-deployment-oriented-compressionSecFeedde2026-08-31T05:32:37.731064+00:00Transformer-Based Autonomous Driving Models and Deployment-Oriented Compression: A Surveyhttps://blog.schoergendorfer.com/post/trajectory-balance-improved-credit-assignment-in-gflownetsSecFeedde2026-08-31T05:32:37.703487+00:00Trajectory balance: Improved credit assignment in GFlowNetshttps://blog.schoergendorfer.com/post/diffusion-models-as-plug-and-play-priorsSecFeedde2026-08-31T05:32:37.676115+00:00Diffusion models as plug-and-play priorshttps://blog.schoergendorfer.com/post/aero-hand-open-a-simulation-ready-tendon-driven-hand-for-dexterous-manipulationSecFeedde2026-08-31T05:17:23.806155+00:00Aero Hand Open: A Simulation-Ready Tendon-Driven Hand for Dexterous Manipulation Learninghttps://blog.schoergendorfer.com/post/learning-between-the-peaks-sharp-asymptotics-for-kernel-ridge-regression-under-pSecFeedde2026-08-31T05:17:23.733098+00:00Learning between the peaks: sharp asymptotics for kernel ridge regression under power-law anisotropyhttps://blog.schoergendorfer.com/post/acquire-repair-preserve-a-diagnosis-guided-post-training-recipe-for-small-modelSecFeedde2026-08-31T05:17:23.708413+00:00Acquire, Repair, Preserve: A Diagnosis-Guided Post-Training Recipe for Small-Model Dialogue Game Agentshttps://blog.schoergendorfer.com/post/generalized-splines-and-gaussian-processesSecFeedde2026-08-31T05:17:23.683113+00:00Generalized Splines and Gaussian Processeshttps://blog.schoergendorfer.com/post/sliding-window-beats-linear-attentionSecFeedde2026-08-31T05:17:23.654162+00:00Sliding-window beats linear attentionhttps://blog.schoergendorfer.com/post/post-training-vlms-for-video-mistake-detectionSecFeedde2026-08-31T05:17:23.626421+00:00Post-Training VLMs for Video Mistake Detectionhttps://blog.schoergendorfer.com/post/timing-aware-repurchase-prediction-for-web-scale-e-commerce-survival-models-forSecFeedde2026-08-31T05:17:23.601837+00:00Timing-Aware Repurchase Prediction for Web-Scale E-Commerce: Survival Models for Multi-Surface Grocery Recommendationhttps://blog.schoergendorfer.com/post/quantum-federated-learning-based-on-bures-uhlmann-geometry-for-heterogeneous-noiSecFeedde2026-08-31T05:17:23.576752+00:00Quantum Federated Learning Based on Bures--Uhlmann Geometry for Heterogeneous Noisy Clientshttps://blog.schoergendorfer.com/post/real-time-monitoring-of-mhd-liquid-metal-flows-with-shallow-recurrent-decodersSecFeedde2026-08-31T05:17:23.541604+00:00Real-Time Monitoring of MHD Liquid Metal Flows with Shallow Recurrent Decodershttps://blog.schoergendorfer.com/post/localizing-global-discrepancies-marginal-contributions-and-contextual-anomaly-deSecFeedde2026-08-31T05:17:23.518952+00:00Localizing Global Discrepancies: Marginal Contributions and Contextual Anomaly Detectionhttps://blog.schoergendorfer.com/post/gracegradient-guided-coreset-selection-for-llm-unlearningSecFeedde2026-08-31T05:17:23.492908+00:00GRACE:Gradient-guided Coreset Selection for LLM Unlearninghttps://blog.schoergendorfer.com/post/banglamed-qa-a-question-answering-system-for-healthcare-support-in-banglaSecFeedde2026-08-31T05:17:23.467598+00:00BanglaMed-QA: A Question Answering System for Healthcare Support in Banglahttps://blog.schoergendorfer.com/post/explainable-diabetic-retinopathy-classification-using-vision-foundation-modelsSecFeedde2026-08-31T05:17:23.394383+00:00Explainable Diabetic Retinopathy Classification Using Vision Foundation Modelshttps://blog.schoergendorfer.com/post/expose-explainable-and-domain-robust-embeddings-from-pathology-vision-foundationSecFeedde2026-08-31T05:17:23.367088+00:00EXPOSE: Explainable and Domain-Robust Embeddings from Pathology Vision Foundation Models using Sparse Autoencodershttps://blog.schoergendorfer.com/post/empowering-local-agriculture-a-deep-learning-powered-web-system-for-identifyingSecFeedde2026-08-31T05:17:23.336241+00:00Empowering Local Agriculture: A Deep Learning-Powered Web System for Identifying Bangladeshi Mango Varietieshttps://blog.schoergendorfer.com/post/under-mattress-temporal-sensing-for-next-day-agitation-risk-scoring-in-dementiaSecFeedde2026-08-31T05:02:43.928593+00:00Under-Mattress Temporal Sensing for Next-Day Agitation Risk Scoring in Dementia Wardshttps://blog.schoergendorfer.com/post/do-medical-vision-models-reason-about-anatomy-probing-the-spatial-inductive-biasSecFeedde2026-08-31T05:02:43.877567+00:00Do Medical Vision Models Reason About Anatomy? Probing the Spatial Inductive Biases of Learned Visual Representationshttps://blog.schoergendorfer.com/post/landau-theory-of-quenched-criticality-in-linear-in-context-learningSecFeedde2026-08-31T05:02:43.854275+00:00Landau theory of quenched criticality in linear in-context learninghttps://blog.schoergendorfer.com/post/emergent-aggregation-from-collective-foragingSecFeedde2026-08-31T05:02:43.824116+00:00Emergent aggregation from collective foraginghttps://blog.schoergendorfer.com/post/characterization-of-request-and-token-energy-costs-for-llm-inference-workloads-oSecFeedde2026-08-31T05:02:43.795702+00:00Characterization of Request and Token Energy Costs for LLM Inference Workloads on GPU Platformshttps://blog.schoergendorfer.com/post/twin-worlds-equivariance-based-abstention-for-evidence-grounded-reasoningSecFeedde2026-08-31T05:02:43.766746+00:00Twin Worlds: Equivariance-Based Abstention for Evidence-Grounded Reasoninghttps://blog.schoergendorfer.com/post/not-to-break-but-to-attest-adversarial-probes-for-privacy-preserving-llm-verificSecFeedde2026-08-31T05:02:43.740680+00:00Not to Break, but to Attest: Adversarial Probes for Privacy-Preserving LLM Verificationhttps://blog.schoergendorfer.com/post/openstamp-a-watermark-for-open-source-language-modelsSecFeedde2026-08-31T05:02:43.713180+00:00OpenStamp: A Watermark for Open-Source Language Modelshttps://blog.schoergendorfer.com/post/what-do-interaction-representations-actually-measure-pre-event-separability-in-wSecFeedde2026-08-31T05:02:43.685556+00:00What Do Interaction Representations Actually Measure? Pre-Event Separability in Weakly-Supervised Violence Detectionhttps://blog.schoergendorfer.com/post/realswe-a-compositional-evaluation-of-coding-agents-under-realistic-user-requestSecFeedde2026-08-31T05:02:43.637787+00:00RealSWE: A Compositional Evaluation of Coding Agents under Realistic User Requestshttps://blog.schoergendorfer.com/post/personalized-and-multi-view-representation-for-federated-cold-start-recommendatiSecFeedde2026-08-31T05:02:43.612542+00:00Personalized and Multi-View Representation for Federated Cold-Start Recommendationhttps://blog.schoergendorfer.com/post/cura-certified-runtime-alarms-for-computer-use-agentsSecFeedde2026-08-31T05:02:43.585847+00:00CURA: Certified Runtime Alarms for Computer-Use Agentshttps://blog.schoergendorfer.com/post/memorization-is-not-extraction-tight-differential-privacy-bounds-and-audit-blindSecFeedde2026-08-31T05:02:43.556754+00:00Memorization Is Not Extraction: Tight Differential-Privacy Bounds and Audit Blind Spotshttps://blog.schoergendorfer.com/post/beyond-procrustes-distances-a-multilinear-gromov-wasserstein-distance-capturingSecFeedde2026-08-31T05:02:43.533085+00:00Beyond Procrustes distances: a multilinear Gromov-Wasserstein distance capturing chiralityhttps://blog.schoergendorfer.com/post/on-the-computational-and-statistical-efficiency-of-the-empirical-maximum-entropySecFeedde2026-08-31T05:02:43.508528+00:00On the Computational and Statistical Efficiency of the Empirical Maximum Entropy on the Mean Methodhttps://blog.schoergendorfer.com/post/cardinal-predicts-cardiovascular-risk-from-non-contrast-cardiac-ctSecFeedde2026-08-31T05:02:43.483636+00:00CARDINAL Predicts Cardiovascular Risk From Non-contrast Cardiac CThttps://blog.schoergendorfer.com/post/quantum-sedonet-spectrally-embedded-quantum-deep-operator-networks-for-partial-dSecFeedde2026-08-31T05:02:43.457353+00:00Quantum SEDONet: Spectrally-Embedded Quantum Deep Operator Networks for Partial Differential Equationshttps://blog.schoergendorfer.com/post/depth-aware-pothole-detection-using-yolo-and-rt-detr-at-the-edgeSecFeedde2026-08-31T05:02:43.430479+00:00Depth-Aware Pothole Detection Using YOLO and RT-DETR at the Edgehttps://blog.schoergendorfer.com/post/tensor-accelerated-eager-multi-resolution-grids-for-evolving-large-scale-substraSecFeedde2026-08-31T04:47:42.745069+00:00Tensor-Accelerated Eager Multi-Resolution Grids for Evolving Large-Scale Substrateshttps://blog.schoergendorfer.com/post/physics-informed-learning-for-the-inverse-problem-in-resonant-ultrasound-spectroSecFeedde2026-08-31T04:47:42.713126+00:00Physics-informed learning for the inverse problem in resonant ultrasound spectroscopyhttps://blog.schoergendorfer.com/post/towards-a-mathematical-theory-of-superpositionSecFeedde2026-08-31T04:47:42.656943+00:00Towards a mathematical theory of superpositionhttps://blog.schoergendorfer.com/post/destroy-me-automatic-artifact-generation-for-histopathology-imagesSecFeedde2026-08-31T04:47:42.603849+00:00Destroy Me: Automatic Artifact Generation for Histopathology Imageshttps://blog.schoergendorfer.com/post/how-do-linear-probes-emerge-a-circuit-tracing-framework-with-concept-targeted-atSecFeedde2026-08-31T04:47:42.579335+00:00How Do Linear Probes Emerge? A Circuit-Tracing Framework with Concept-Targeted Attributionhttps://blog.schoergendorfer.com/post/multiscale-community-based-fingerprinting-of-signed-functional-networksSecFeedde2026-08-31T04:47:42.524397+00:00Multiscale Community-Based Fingerprinting of Signed Functional Networkshttps://blog.schoergendorfer.com/post/effectiveness-of-iot-and-deep-learning-for-detection-and-severity-assessment-ofSecFeedde2026-08-31T04:47:42.494305+00:00Effectiveness of IoT and Deep Learning for Detection and Severity Assessment of Postelectrotermes militaris in Tea Plantationshttps://blog.schoergendorfer.com/post/hypothesize-evaluate-refine-a-scientific-agent-for-pde-discovery-with-unknown-spSecFeedde2026-08-31T04:47:42.461798+00:00Hypothesize, Evaluate, Refine: A Scientific Agent for PDE Discovery with Unknown Spatial Coefficient Fieldshttps://blog.schoergendorfer.com/post/scirec-diagnostic-evaluation-of-multimodal-multi-turn-relational-reasoning-withSecFeedde2026-08-31T04:47:42.432548+00:00SciReC: Diagnostic Evaluation of Multimodal, Multi-Turn Relational Reasoning with Adaptive Interactionhttps://blog.schoergendorfer.com/post/qgpinns-a-physics-informed-neural-network-framework-for-nonlocal-differential-eqSecFeedde2026-08-31T04:47:42.372259+00:00QGPINNs: A Physics-Informed Neural Network Framework for Nonlocal Differential Equations on Quantum Graphshttps://blog.schoergendorfer.com/post/blog-survey-of-optimizersSecFeedde2026-08-31T04:47:42.339449+00:00Blog: Survey of Optimizershttps://blog.schoergendorfer.com/post/advancing-interaction-sensitive-feature-selection-novel-relief-based-algorithmsSecFeedde2026-08-31T04:47:42.308682+00:00Advancing Interaction-Sensitive Feature Selection: Novel Relief-Based Algorithms, Expanded Comparisons, and Recommendations for Biomedical Data Mininghttps://blog.schoergendorfer.com/post/darts-decoder-aware-representation-tuning-via-surgery-for-model-mergingSecFeedde2026-08-31T04:47:42.274110+00:00DARTS: Decoder-Aware Representation Tuning via Surgery for Model Merginghttps://blog.schoergendorfer.com/post/an-enclosed-mode-is-a-gauge-choice-topology-relative-to-reach-in-certified-codeSecFeedde2026-08-31T04:47:42.250229+00:00An Enclosed Mode Is a Gauge Choice: Topology Relative to Reach in Certified Code World Modelshttps://blog.schoergendorfer.com/post/replicant-learning-policies-for-evading-and-hardening-malware-detectorsSecFeedde2026-08-31T04:47:42.219019+00:00REPLICANT: Learning Policies for Evading and Hardening Malware Detectorshttps://blog.schoergendorfer.com/post/how-proper-scoring-rules-shape-llm-forecastingSecFeedde2026-08-31T04:47:42.189241+00:00How Proper Scoring Rules Shape LLM Forecastinghttps://blog.schoergendorfer.com/post/curvature-conditioned-multiscale-momentum-with-sphere-constraints-for-llm-pretraSecFeedde2026-08-31T04:32:26.285237+00:00Curvature-Conditioned Multiscale Momentum with Sphere Constraints for LLM Pretraininghttps://blog.schoergendorfer.com/post/euclidean-fourier-neural-operatorsSecFeedde2026-08-31T04:32:26.259150+00:00Euclidean Fourier Neural Operatorshttps://blog.schoergendorfer.com/post/symbollm-fe-llm-accelerated-symbolic-regression-for-automated-feature-engineerinSecFeedde2026-08-31T04:32:26.232086+00:00SymboLLM-FE: LLM-Accelerated Symbolic Regression for Automated Feature Engineering on Tabular Datahttps://blog.schoergendorfer.com/post/deriving-scaling-laws-for-openeurollm-models-learning-rate-batch-size-and-lossSecFeedde2026-08-31T04:32:26.205387+00:00Deriving Scaling Laws for OpenEuroLLM Models: Learning Rate, Batch Size and Losshttps://blog.schoergendorfer.com/post/vista-verifier-informed-student-to-teacher-adaptation-for-on-policy-self-distillSecFeedde2026-08-31T04:32:26.177210+00:00VISTA: Verifier-Informed Student-to-Teacher Adaptation for On-Policy Self-Distillationhttps://blog.schoergendorfer.com/post/parser-states-already-know-structure-conditioned-kv-persistence-for-structured-gSecFeedde2026-08-31T04:32:26.145775+00:00Parser States Already Know: Structure-Conditioned KV Persistence for Structured Generationhttps://blog.schoergendorfer.com/post/learning-to-transfer-across-modes-towards-unified-urban-mobility-forecastingSecFeedde2026-08-31T04:32:26.096060+00:00Learning to Transfer Across Modes: Towards Unified Urban Mobility Forecastinghttps://blog.schoergendorfer.com/post/residual-guided-randomized-neural-networksSecFeedde2026-08-31T04:32:26.066092+00:00Residual-Guided Randomized Neural Networkshttps://blog.schoergendorfer.com/post/sinkslot-sinkhorn-via-sparse-lifted-optimal-transportSecFeedde2026-08-31T04:32:26.035080+00:00SinkSLOT: Sinkhorn via Sparse Lifted Optimal Transporthttps://blog.schoergendorfer.com/post/spectral-features-dominate-bcg-respiratory-event-detection-a-large-scale-patientSecFeedde2026-08-31T04:32:26.010371+00:00Spectral Features Dominate BCG Respiratory-Event Detection: A Large-Scale Patient-Independent Comparison of Feature Groups in Sleep Apnea Patientshttps://blog.schoergendorfer.com/post/efficient-online-continual-foundation-model-fine-tuning-for-predictive-process-mSecFeedde2026-08-31T04:32:25.984613+00:00Efficient Online Continual Foundation Model Fine-Tuning for Predictive Process Monitoringhttps://blog.schoergendorfer.com/post/d-taia-domain-aware-llm-adaptation-for-multi-task-predictive-process-monitoringSecFeedde2026-08-31T04:32:25.956887+00:00D-TAIA: Domain-Aware LLM Adaptation for Multi-Task Predictive Process Monitoringhttps://blog.schoergendorfer.com/post/generalized-context-in-cross-attention-for-transfer-learning-of-disjoint-tabularSecFeedde2026-08-31T04:32:25.930519+00:00Generalized Context in Cross Attention for Transfer Learning of Disjoint Tabular Datahttps://blog.schoergendorfer.com/post/performative-privacy-when-differential-privacy-maximizes-utilitySecFeedde2026-08-31T04:32:25.903692+00:00Performative Privacy: When Differential Privacy Maximizes Utilityhttps://blog.schoergendorfer.com/post/biologically-inspired-mechanisms-for-facilitating-grokking-in-multilayer-perceptSecFeedde2026-08-31T04:32:25.854123+00:00Biologically Inspired Mechanisms for Facilitating Grokking in Multilayer Perceptronshttps://blog.schoergendorfer.com/post/harts-efficient-agentic-reinforcement-learning-for-hybrid-attention-models-overSecFeedde2026-08-31T04:32:25.826344+00:00HARTS: Efficient Agentic Reinforcement Learning for Hybrid-Attention Models over Arbitrary Rollout Treeshttps://blog.schoergendorfer.com/post/the-approximation-rank-of-softmax-attention-sharp-geometric-laws-and-robust-inteSecFeedde2026-08-31T04:32:25.797258+00:00The Approximation Rank of Softmax Attention: Sharp Geometric Laws and Robust Interaction Dimensionhttps://blog.schoergendorfer.com/post/conditional-diffusion-models-for-energy-efficient-drivingSecFeedde2026-08-31T04:32:25.763778+00:00Conditional Diffusion Models for Energy-Efficient Drivinghttps://blog.schoergendorfer.com/post/learning-to-difference-adaptive-reversible-differencing-adardiff-for-time-seriesSecFeedde2026-08-31T04:18:49.389547+00:00Learning to Difference: Adaptive Reversible Differencing (AdaRDiff) for Time Series Forecastinghttps://blog.schoergendorfer.com/post/vict-verifier-instrumented-credit-tracing-for-long-horizon-llm-agent-reinforcemeSecFeedde2026-08-31T04:18:49.360373+00:00VICT: Verifier-Instrumented Credit Tracing for Long-Horizon LLM Agent Reinforcement Learninghttps://blog.schoergendorfer.com/post/generalized-gibbs-ensemble-weighting-for-forecast-combinationSecFeedde2026-08-31T04:18:49.331549+00:00Generalized Gibbs Ensemble Weighting for Forecast Combinationhttps://blog.schoergendorfer.com/post/comparing-classical-and-quantum-machine-learning-for-regression-in-high-energy-pSecFeedde2026-08-31T04:18:49.302419+00:00Comparing Classical and Quantum Machine Learning for Regression in High Energy Physics Collision Datahttps://blog.schoergendorfer.com/post/explainable-uncertainty-estimation-for-reliable-medical-aiSecFeedde2026-08-31T04:18:49.273928+00:00Explainable Uncertainty Estimation for Reliable Medical AIhttps://blog.schoergendorfer.com/post/when-can-conditional-flow-matching-replace-pointwise-negative-log-likelihoodSecFeedde2026-08-31T04:18:49.245059+00:00When Can Conditional Flow Matching Replace Pointwise Negative Log-Likelihood?https://blog.schoergendorfer.com/post/exact-risk-ratios-for-weighted-data-selection-in-linear-regressionSecFeedde2026-08-31T04:18:49.220822+00:00Exact Risk Ratios for Weighted Data Selection in Linear Regressionhttps://blog.schoergendorfer.com/post/is-monte-carlo-tree-search-just-every-visit-monte-carlo-controlSecFeedde2026-08-31T04:18:49.190160+00:00Is Monte Carlo Tree Search Just Every-Visit Monte Carlo Control?https://blog.schoergendorfer.com/post/a-method-for-layer-bit-width-allocation-in-llm-quantization-via-performance-maxiSecFeedde2026-08-31T04:18:49.162606+00:00A Method for Layer Bit-Width Allocation in LLM Quantization via Performance Maximization Under a Quality-Degradation Constrainthttps://blog.schoergendorfer.com/post/phymamba-physics-modulated-mamba-for-robust-battery-health-prognosticsSecFeedde2026-08-31T04:18:49.133729+00:00PhyMamba: Physics-Modulated Mamba for Robust Battery Health Prognosticshttps://blog.schoergendorfer.com/post/temporal-memory-aware-online-test-time-adaptation-on-dynamic-graphsSecFeedde2026-08-31T04:18:49.102937+00:00Temporal Memory-Aware Online Test-Time Adaptation on Dynamic Graphshttps://blog.schoergendorfer.com/post/ti2ps-a-topology-informed-inverse-design-framework-for-stochastic-multicellularSecFeedde2026-08-31T04:18:49.071609+00:00TI$^2$PS: A Topology-Informed Inverse Design Framework for Stochastic Multicellular Pattern Formationhttps://blog.schoergendorfer.com/post/tacit-switch-cost-aware-model-escalation-for-llm-agents-from-censored-supervisioSecFeedde2026-08-31T04:18:49.042531+00:00TACIT-Switch: Cost-Aware Model Escalation for LLM Agents from Censored Supervisionhttps://blog.schoergendorfer.com/post/beyond-pairwise-graphs-in-science-hypergraph-adaptive-wavelet-operators-for-paraSecFeedde2026-08-31T04:18:48.986570+00:00Beyond Pairwise Graphs in Science: Hypergraph Adaptive Wavelet Operators for Parametric PDEshttps://blog.schoergendorfer.com/post/somtab-set-order-mamba-for-efficient-tabular-in-context-learningSecFeedde2026-08-31T04:18:48.960224+00:00SOMTab: Set-Order Mamba for Efficient Tabular In-Context Learninghttps://blog.schoergendorfer.com/post/fedehr-agents-federated-agentic-optimization-for-automated-ehr-modelingSecFeedde2026-08-31T04:18:48.929539+00:00FedEHR-Agents: Federated Agentic Optimization for Automated EHR Modelinghttps://blog.schoergendorfer.com/post/actionable-cbfi-integrating-structural-decomposition-and-causal-counterfactual-rSecFeedde2026-08-31T04:18:48.903697+00:00Actionable CBFI: Integrating Structural Decomposition and Causal Counterfactual Recourse for Tabular Machine Learninghttps://blog.schoergendorfer.com/post/initialization-is-critical-advancing-federated-short-term-load-forecasting-underSecFeedde2026-08-31T04:18:48.853276+00:00Initialization Is Critical: Advancing Federated Short-Term Load Forecasting under Load Heterogeneity via Model Initializationhttps://blog.schoergendorfer.com/post/beyond-search-imitation-prior-directed-exploration-for-searchless-chessSecFeedde2026-08-31T04:02:24.001395+00:00Beyond Search-Imitation: Prior-Directed Exploration for Searchless Chesshttps://blog.schoergendorfer.com/post/the-calls-are-coming-from-inside-the-model-investigating-probe-based-detection-oSecFeedde2026-08-31T04:02:23.972490+00:00The Calls are Coming from Inside the Model: Investigating Probe-based Detection of Tool-Calling Errors in LLMshttps://blog.schoergendorfer.com/post/leveraging-a-foundation-model-for-the-eeg-based-diagnosis-of-alzheimers-diseaseSecFeedde2026-08-31T04:02:23.924383+00:00Leveraging a Foundation Model for the EEG-Based Diagnosis of Alzheimer's Diseasehttps://blog.schoergendorfer.com/post/beyond-non-iid-learner-client-distribution-mismatch-in-federated-learningSecFeedde2026-08-31T04:02:23.899380+00:00Beyond Non-IID: Learner--Client Distribution Mismatch in Federated Learninghttps://blog.schoergendorfer.com/post/riskblend-a-multi-signal-framework-for-test-input-prioritization-in-machine-learSecFeedde2026-08-31T04:02:23.870231+00:00RiskBlend: A Multi-Signal Framework for Test Input Prioritization in Machine Learning Regression Testinghttps://blog.schoergendorfer.com/post/dart-fl-burst-aware-multitask-federated-learning-under-dynamic-inference-demandSecFeedde2026-08-31T04:02:23.839584+00:00DART-FL: Burst-Aware Multitask Federated Learning under Dynamic Inference Demand at the Edgehttps://blog.schoergendorfer.com/post/safestep-an-interactive-demonstration-of-semantic-communication-for-pedestrian-sSecFeedde2026-08-31T04:02:23.814747+00:00SafeStep: An Interactive Demonstration of Semantic Communication for Pedestrian Safety Monitoringhttps://blog.schoergendorfer.com/post/segbench-gc-testing-segmentation-invariance-in-multi-step-offline-goal-conditionSecFeedde2026-08-31T04:02:23.783810+00:00SegBench-GC: Testing Segmentation Invariance in Multi-Step Offline Goal-Conditioned Reinforcement Learninghttps://blog.schoergendorfer.com/post/unsupervised-continual-learning-with-growing-self-organizing-maps-and-syntheticSecFeedde2026-08-31T04:02:23.758467+00:00Unsupervised Continual Learning with Growing Self-Organizing Maps and Synthetic Replayhttps://blog.schoergendorfer.com/post/curvature-aware-radius-shrinkage-for-adaptive-nearest-neighbor-classificationSecFeedde2026-08-31T04:02:23.730676+00:00Curvature-Aware Radius Shrinkage for Adaptive Nearest Neighbor Classificationhttps://blog.schoergendorfer.com/post/more-data-cannot-break-a-symmetry-identifiability-by-designSecFeedde2026-08-31T04:02:23.705082+00:00More Data Cannot Break a Symmetry: Identifiability by Designhttps://blog.schoergendorfer.com/post/self-explainable-multi-label-graph-neural-network-for-correlated-evidence-attribSecFeedde2026-08-31T04:02:23.678769+00:00Self-Explainable Multi-Label Graph Neural Network for Correlated Evidence Attributionhttps://blog.schoergendorfer.com/post/dandelion-a-spherical-flower-for-neural-simulation-of-planetary-dynamicsSecFeedde2026-08-31T04:02:23.652233+00:00Dandelion: A Spherical Flower for Neural Simulation of Planetary Dynamicshttps://blog.schoergendorfer.com/post/when-muon-meets-task-interference-a-spectral-perspective-on-continual-learning-aSecFeedde2026-08-31T04:02:23.624488+00:00When Muon Meets Task Interference: A Spectral Perspective on Continual Learning and Model Merginghttps://blog.schoergendorfer.com/post/damp-decay-aware-mixed-precision-recurrent-state-quantizationSecFeedde2026-08-31T04:02:23.569593+00:00DAMP: Decay-Aware Mixed-Precision Recurrent-State Quantizationhttps://blog.schoergendorfer.com/post/quantization-triggered-backdoors-in-language-models-cross-quantizer-transferabilSecFeedde2026-08-31T04:02:23.539821+00:00Quantization-Triggered Backdoors in Language Models: Cross-Quantizer Transferability and the Validation--Deployment Gaphttps://blog.schoergendorfer.com/post/marginal-coverage-credit-reduces-redundant-exploration-in-parallel-state-entropySecFeedde2026-08-31T04:02:23.485630+00:00Marginal Coverage Credit Reduces Redundant Exploration in Parallel State-Entropy Optimizationhttps://blog.schoergendorfer.com/post/understanding-chatgpt-workSecFeedde2026-08-31T00:17:29.267438+00:00Understanding ChatGPT Workhttps://blog.schoergendorfer.com/post/lowest-latency-inference-apis-for-voice-and-realtime-agents-a-time-to-first-tokeSecFeedde2026-08-30T21:32:46.298939+00:00Lowest-Latency Inference APIs for Voice and Realtime Agents: A Time to First Token TTFT-First Benchmarkhttps://blog.schoergendorfer.com/post/google-ai-introduces-envharness-a-programmable-layer-that-turns-static-agent-envSecFeedde2026-08-30T20:33:38.497725+00:00Google AI Introduces EnvHarness: A Programmable Layer That Turns Static Agent Environments Into Adaptive Training Worldshttps://blog.schoergendorfer.com/post/openai-leaving-cursor-developers-have-to-be-prepared-to-adapt-when-it-happensSecFeedde2026-08-30T17:18:10.091790+00:00OpenAI leaving Cursor: “Developers have to be prepared to adapt when it happens.”https://blog.schoergendorfer.com/post/ai-agents-are-making-retrieval-engineering-a-core-engineering-disciplineSecFeedde2026-08-30T16:18:53.148209+00:00AI agents are making retrieval engineering a core engineering disciplinehttps://blog.schoergendorfer.com/post/your-ai-agent-is-only-as-good-as-the-harness-around-itSecFeedde2026-08-30T15:18:37.405373+00:00Your AI agent is only as good as the harness around ithttps://blog.schoergendorfer.com/post/anthropic-opens-a-research-preview-of-the-model-hardware-standard-mhs-a-shared-sSecFeedde2026-08-30T06:17:22.138434+00:00Anthropic Opens a Research Preview of the Model Hardware Standard (MHS): A Shared Specification for AI Agents to Safely Operate Physical Deviceshttps://blog.schoergendorfer.com/post/meet-code-as-world-an-agentic-loop-that-rewrites-real-videos-into-executable-mujSecFeedde2026-08-30T01:48:26.242273+00:00Meet ‘Code-as-World’: An Agentic Loop That Rewrites Real Videos Into Executable MuJoCo Physics Programshttps://blog.schoergendorfer.com/post/introducing-hy4-previewSecFeedde2026-08-30T00:17:20.747905+00:00Introducing Hy4 Previewhttps://blog.schoergendorfer.com/post/building-custom-batched-ensemble-weather-forecasting-with-nvidia-earth2studioSecFeedde2026-08-29T19:03:26.463743+00:00Building Custom Batched Ensemble Weather Forecasting with NVIDIA Earth2Studiohttps://blog.schoergendorfer.com/post/deadlock-ransomware-breaking-down-a-rust-based-encryptor-with-decentralized-reco-2SecFeedde2026-08-29T15:48:14.570171+00:00DeadLock ransomware: Breaking down a Rust-based encryptor with decentralized recovery infrastructurehttps://blog.schoergendorfer.com/post/hunting-macsync-stealer-infrastructure-through-behavioral-pivots-2SecFeedde2026-08-29T15:48:14.523306+00:00Hunting MacSync Stealer infrastructure through behavioral pivotshttps://blog.schoergendorfer.com/post/when-ai-infrastructure-becomes-the-target-securing-gateways-and-control-points-2SecFeedde2026-08-29T15:48:14.483411+00:00When AI infrastructure becomes the target: Securing gateways and control pointshttps://blog.schoergendorfer.com/post/your-container-runs-everything-around-it-shouldnt-be-your-problemSecFeedde2026-08-29T15:17:50.257164+00:00Your container runs. Everything around it shouldn’t be your problem.https://blog.schoergendorfer.com/post/google-ai-releases-gemini-omni-11-flash-40-second-scene-extension-firstlast-framSecFeedde2026-08-29T14:47:40.136537+00:00Google AI Releases Gemini Omni 1.1 Flash: 40-Second Scene Extension, First/Last Frame Control, and 4K Upscalinghttps://blog.schoergendorfer.com/post/commits-on-github-have-doubled-in-four-months-verification-capacity-has-notSecFeedde2026-08-29T14:02:59.106180+00:00Commits on GitHub have doubled in four months. Verification capacity has not.