CORTEXA
← Browse
arxivcs.LGcs.AIcs.CL2026-07-16

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

Weiwen Xu, Jia Liu, Hou Pong Chan, Long Li, Deng Cai, Min Chen, Hao Zhang

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.

View free PDFSource page

Related papers

arxivcs.AIcs.CLcs.LG2026-07-21

Measuring Reward-Seeking via Contrastive Belief Updates

Axel Højmark, Jérémy Scheurer, Evgenia Nitishinskaya, Felix Hofstätter, Jason Wolfe, Theodore Ehrenborg, et al.

Language models trained with reinforcement learning may learn to optimize the grader's judgment rather than the intended objective. This "reward-seeking" is difficult to measure because a model that pursues the grader's judgment and one that pursues the intended objective behave…

View free PDFSource page
arxivcs.CLcs.AIcs.LG2026-07-16

Mask-Aware Policy Gradients for Diffusion Language Models

Haran Raajesh, Kulin Shah, Adam Klivans, Philipp Krähenbühl

Reinforcement learning has proven effective for improving reasoning in large language models, but extending it to Masked Diffusion Language Models (MDLMs) remains challenging due to the intractability of the log-likelihood estimation. Existing approaches approximate this log-like…

View free PDFSource page
arxivcs.AIcs.CLcs.LG2026-07-19

Otap:Structure-Aware Optimal Transport for Evaluating Planning and Execution in Agent Trajectories

Babak Barazandeh, Subhabrata Majumdar, George Michailidis

Large language model agents solve tasks by generating trajectories that interleave planning, tool calls, and intermediate results. Current evaluation metrics reduce such a trajectory to a binary success flag or compare it against a reference by exact matching. A success flag cann…

View free PDFSource page
arxivcs.LGcs.AIcs.CL2026-07-24

Cross-Tokenizer On-Policy Distillation via Byte-Prefix Marginalization

Hao Wang, Kun Yuan, Wenlin Zhong, Minglei Zhang, Han Xiao, Ming Sun, et al.

Open-weight language models from different families exhibit complementary capabilities, motivating their consolidation into a compact student through on-policy distillation (OPD). However, full-vocabulary OPD typically assumes a shared tokenizer, while existing cross-tokenizer me…

View free PDFSource page