paper-with-me

홈 › Papers

PRPO: Aligning Process Reward with Outcome Reward in Policy Optimization

2026-01-12 · Ruiyi Ding, Yongxuan Lv, Xianhui Meng, Jiahe Song, Chao Wang, Chen Jiang, Yuan Cheng arxiv

Policy optimization for large language models often suffers from sparse reward signals in multi-step reasoning tasks. Critic-free methods like GRPO assign a single normalized outcome reward to all tokens, providing limited guidance for intermediate reasoning . While Process Reward Models (PRMs) offer dense feedback, they risk premature collapse when used alone, as early low-reward tokens can drive policies toward truncated outputs. We introduce Process Relative Policy Optimization (PRPO), which combines outcome reliability with process-level guidance in a critic-free framework. PRPO segments reasoning sequences based on semantic clues, normalizes PRM scores into token-level advantages, and aligns their distribution with outcome advantages through location-parameter shift. On MATH500, PRPO improves Qwen2.5-Math-1.5B accuracy from 61.2% to 64.4% over GRPO using only eight rollouts and no value network, demonstrating efficient fine-grained credit assignment within critic-free optimization. Code is available at: https://github.com/SchumiDing/srpocode

📄 PDF Abstract BibTeX arXiv:2601.07182

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Mitigating Preference Hacking in Policy Optimization with Pessimism

2025-03-10 · Dhawal Gupta, Adam Fisch, Christoph Dann, Alekh Agarwal

This work tackles the problem of overoptimization in reinforcement learning from human feedback (RLHF), a prevalent technique for aligning models with human preferences. RLHF relies on reward or preference models trained…

Document Summarization

PRPO: Perception-Reinforced Policy Optimization via Token-Level Dynamic Advantage Reshaping

2026-06-07 · Qiming Li, Tianlun Li, Xiaolong Cheng, Hangyu Li 외 arxiv

Reinforcement Learning with Verifiable Rewards (RLVR) has become an effective paradigm for improving the reasoning capability of Large Vision-Language Models (LVLMs). However, existing RLVR methods primarily rely on traj…

Reinforcement LearningMultimodal Reasoning

Euphonium: Steering Video Flow Matching via Process Reward Gradient Guided Stochastic Dynamics

2026-02-04 · Ruizhe Zhong, Jiesong Lian, Xiaoyue Mi, Zixiang Zhou 외 arxiv

While online Reinforcement Learning has emerged as a crucial technique for aligning flow matching models with human preferences, current approaches are hindered by inefficient exploration during training rollouts. Relyin…

Text-to-Video GenerationReinforcement Learning

Chart Deep Research in LVLMs via Parallel Relative Policy Optimization

2026-03-03 · Jiajin Tang, Gaoyang, Wenjie Wang, Sibei Yang 외 arxiv

With the rapid advancement of data science, charts have evolved from simple numerical presentation tools to essential instruments for insight discovery and decision-making support. However, current chart data intelligenc…

Reward Collapse in Aligning Large Language Models

2023-05-28 · Ziang Song, Tianle Cai, Jason D. Lee, Weijie J. Su

The extraordinary capabilities of large language models (LLMs) such as ChatGPT and GPT-4 are in part unleashed by aligning them with reward models that are trained on human preferences, which are often represented as ran…