paper-with-me

Papers

Beyond Entropy: Correctness-Aware Advantage Shaping via Contrastive Policy Optimization

2026-07-16 · Weiwen Xu, Jia Liu, Hou Pong Chan, Long Li, Deng Cai, Min Chen, Hao Zhang arxiv

Reinforcement learning with verifiable rewards (RLVR) commonly uses entropy for advantage shaping. However, entropy cannot distinguish useful uncertainty from detrimental confusion, limiting its effectiveness as a correctness signal. We propose Contrastive Policy Optimization (CPO), which uses token-level contrastive disagreement between reference-guided and vanilla generation distributions for correctness-aware advantage shaping. Both theoretical and empirical results show that this disagreement reliably indicates token-level correctness. We further show that On-policy Distillation is a special case of CPO, where the posterior distribution is instantiated by an external teacher model. CPO also resolves the zero-advantage problem. Experiments on in-domain and out-of-domain benchmarks demonstrate that CPO substantially outperforms entropy-based RLVR methods while maintaining strong generalization. Further analysis shows that correct and incorrect responses naturally support exploration and exploitation respectively, and balancing both leads to the best performance.

📄 PDF Abstract BibTeX arXiv:2607.14614

Code (2)

Tavish9/awesome-daily-AI-arxiv ★ 111
yingyingxia666/awesome-agentic ★ 23

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

Beyond High-Entropy Exploration: Correctness-Aware Low-Entropy Segment-Based Advantage Shaping for Reasoning LLMs

2025-11-30 · Xinzhu Chen, Xuesheng Li, Zhongxiang Sun, Weijie Yu arxiv

Reinforcement Learning with Verifiable Rewards (RLVR) has become a central approach for improving the reasoning ability of large language models. Recent work studies RLVR through token entropy, arguing that high-entropy …

Reinforcement Learning

Unlocking Exploration in RLVR: Uncertainty-aware Advantage Shaping for Deeper Reasoning

2025-10-12 · Can Xie, Ruotong Pan, Xiangyu Wu, Yunfei Zhang 외 arxiv

Reinforcement Learning with Verifiable Rewards (RLVR) has shown significant promise for enhancing the reasoning capabilities of large language models (LLMs). However, prevailing algorithms like GRPO broadcast a uniform a…

Reinforcement LearningMathematical Reasoning

Stable Adaptive Thinking via Advantage Shaping and Length-Aware Gradient Regulation

2026-02-26 · Zihang Xu, Haozhi Xie, Ziqi Miao, Wuxuan Gong 외 arxiv

Large reasoning models (LRMs) achieve strong performance through extended reasoning traces, but they often exhibit overthinking behavior for low-complexity queries. Existing efforts to mitigate this issue are fundamental…

Reinforcement Learning

IAPO: Information-Aware Policy Optimization for Token-Efficient Reasoning

2026-02-22 · Yinhan He, Yaochen Zhu, Mingjia Shi, Wendy Zheng 외 arxiv

Large language models increasingly rely on long chains of thought to improve accuracy, yet such gains come with substantial inference-time costs. We revisit token-efficient post-training and argue that existing sequence-…

AdaTIR: Adaptive Tool-Integrated Reasoning via Difficulty-Aware Policy Optimization

2026-01-21 · Zhaiyu Fang, Ruipeng Sun arxiv

Tool-Integrated Reasoning (TIR) has significantly enhanced the capabilities of Large Language Models (LLMs), yet current agents tend to exhibit cognitive offloading, redundantly invoking external tools even for simple ta…