Researcher — paper.dinq.me

作者位次 #6/8ICML 2026Poster· Applications / Everything Else

ProRL: Effective Reinforcement Learning for Proactive Recommendation via Rectified Policy Gradient Estimation

Hongru Hou · Tiehua Mei · Denghui Geng · Jinhui Huang · Ao Xu · Hengrui Chen · Jiaqing Liang · Deqing Yang

强化学习与后训练OpenReview

作者位次 #6/8ACL 2026Findings· Findings of ACL

Good Reasoning Makes Good Demonstrations: Implicit Reasoning Quality Supervision via In-Context Reinforcement Learning

Tiehua Mei · Minxuan Lv · Leiyu Pan · Zhenpeng Su · Hongru Hou · Hengrui Chen · Ao Xu · Deqing Yang

摘要

Reinforcement Learning with Verifiable Rewards (RLVR) improves reasoning in large language models but treats all correct solutions equally, potentially reinforcing flawed traces that arrive at correct answers by chance. We observe that better reasoning makes better demonstrations : high-quality s...

强化学习与后训练推理评测与基准OpenReview 项目页

Hengrui Chen

获取完整研究档案

联系方式

2026 顶会论文

ProRL: Effective Reinforcement Learning for Proactive Recommendation via Rectified Policy Gradient Estimation

Good Reasoning Makes Good Demonstrations: Implicit Reasoning Quality Supervision via In-Context Reinforcement Learning

常见合作者