Surprising Success, Repeated Failure: Entropy-Guided Credit Assignment for Exploration in LLM Reasoning
cs.LG, cs.AI, cs.CL
Submitted: 2026-09-27
Updated: 2026-09-27
Code: https://github.com/huggingface/math-verify
Project page: https://eapo-explore.github.io
Terminology
Sources
- TreeAdv: Tree-Structured Advantage Redistribution for Group-Based RL
- Evaluating Large Language Models Trained on Code
- Group Entropy-Controlled Policy Optimization
- The Entropy Mechanism of Reinforcement Learning for Reasoning Language Models
- Does On-Policy Distillation Really Distill? From Noisy Teacher to Self-Improvement
- Where Hindsight Credit Can Reside: A Signed-Capacity View of Token Updates in RLVR
- Understanding Reasoning in LLMs through Strategic Information Allocation under Uncertainty
- OPPO: Bayesian Value Recursion for Token-Level Credit Assignment in LLM Reasoning
- AlphaEvolve: A coding agent for scientific and algorithmic discovery
- Olmo 3
- GRAIL: Gradient-Reweighted Advantages for Reinforcement Learning with Verifiable Rewards
- Qwen3 Technical Report
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- Self-Distilled RLVR
- Entropy Polarity in Reinforcement Fine-Tuning: Direction, Asymmetry, and Control
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks