Artificial Intelligence · 26.08.2026, 06:17 UTC
FARCA: Fact-Aligned Reliability-Aware Credit Assignment for Reinforcement Learning with Factual Supervision
| Schweregrad | info |
|---|---|
| Kategorie | Artificial Intelligence |
| Quelle | arXiv cs.AI ↗ |
| Veröffentlicht | 26.08.2026 UTC |
Sicherheitsmeldung mit Schweregrad noch nicht bewertet. Technische Details im Tab „Originaltext“; empfohlene Schritte in der Checkliste.
arXiv:2608.24350v1 Announce Type: cross Abstract: To reduce the hallucination risk caused by outcome-driven rewards in large language models trained through reinforcement learning with verifiable rewards, existing mitigation approaches introduce process-level factual supervision. However, due to coarse-grained aggregation of factual signals and the lack of reliability assessment for these signals, they create a mismatch between fact verification and policy updates. We term this noisy factual credit assignment and decompose it into two aspects: credit localization ambiguity and credit reliability ambiguity. To address these issues, we propose FARCA (Fact-Aligned Reliability-Aware Credit Assignment), a policy optimization framework that transforms factual supervision into localized, reliability-weighted token-level training signals. FARCA achieves fine-grained credit localization by aligning the granularity of fact verification with that of policy updates. It further introduces counterfactual evidence attribution, which uses the dependence of a factual judgment on key evidence as an empirical proxy for verification reliability to compute reliability weights. These weights modulate factual rewards and local policy advantages, reducing the influence of potentially unreliable signals on policy optimization. Experiments across different models and multiple factual reasoning benchmarks show that FARCA significantly improves model factuality while preserving general reasoning capabilities.
Maßnahmen
⬇ Als MarkdownVerwandte Beiträge
- info msData: A Millisecond-Resolution Network Dataset for Advancing Time Series Foundation Models
- info ADVERSA: Measuring Multi-Turn Guardrail Degradation and Judge Reliability in Large Language Models
- info EstLLM: Enhancing Estonian Capabilities in Multilingual LLMs via Continued Pretraining and Post-Training
- info ST-Lite: Training-Free KV Cache Compression with Spatio-Trajectory Guidance for Long-Horizon GUI Agents