2 papers
cs.LG2026
Physics-Guided Policy Optimization with Self-Distillation
Ke Wang, Yuning Wu, Haoran Liu +3
Self-distilled policy optimization (SDPO) has become a popular paradigm for LLM post-training, where a model learns from its own predictions conditioned on privileged information.…
cs.LG2026
Hindsight-Anchored Policy Optimization: Turning Failure into Feedback in Sparse Reward Settings
Yuning Wu, Ke Wang, Devin Chen +1
Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a promising paradigm for post-training reasoning models. However, group-based methods such as Group Relative Po…