paper-with-me

홈 › Papers

GVPO: Group Variance Policy Optimization for Large Language Model Post-Training

2025-04-28 · Kaichen Zhang, Yuzhong Hong, Junwei Bao, Hongfei Jiang, Yang song, Dingqian Hong, Hui Xiong

Post-training plays a crucial role in refining and aligning large language models to meet specific tasks and human preferences. While recent advancements in post-training techniques, such as Group Relative Policy Optimization (GRPO), leverage increased sampling with relative reward scoring to achieve superior performance, these methods often suffer from training instability that limits their practical adoption. To address this challenge, we present Group Variance Policy Optimization (GVPO). GVPO incorporates the analytical solution to KL-constrained reward maximization directly into its gradient weights, ensuring alignment with the optimal policy. The method provides intuitive physical interpretations: its gradient mirrors the mean squared error between the central distance of implicit rewards and that of actual rewards. GVPO offers two key advantages: (1) it guarantees a unique optimal solution, exactly the KL-constrained reward maximization objective, (2) it supports flexible sampling distributions that avoids on-policy and importance sampling limitations. By unifying theoretical guarantees with practical adaptability, GVPO establishes a new paradigm for reliable and versatile LLM post-training.

📄 PDF Abstract BibTeX arXiv:2504.19599

Code (0)

등록된 구현이 없습니다.

Tasks

Language ModelingLanguage ModellingLarge Language Model

Similar Papers 제목 키워드 기반

LongVPO: From Anchored Cues to Self-Reasoning for Long-Form Video Preference Optimization

2026-02-02 · Zhenpeng Huang, Jiaqi Li, Zihan Jia, Xinhao Li 외 arxiv

We present LongVPO, a novel two-stage Direct Preference Optimization framework that enables short-context vision-language models to robustly understand ultra-long videos without any long-video annotations. In Stage 1, we…

BiasGRPO: Stabilizing Bias Mitigation in High-Variance Reward Landscapes via Group-Relative Policy Optimization

2026-06-03 · Saket Reddy, Ke Yang, ChengXiang Zhai arxiv

Mitigating social bias in Large Language Models (LLMs) presents a distinct alignment challenge: unlike verifiable tasks, bias lacks a single ground truth, creating a high-variance, subjective reward landscape. Previous p…

Adaptive Group Policy Optimization: Towards Stable Training and Token-Efficient Reasoning

2025-03-20 · Chen Li, Nazhou Liu, Kai Yang

Since DeepSeek-R1 popularized, Group Relative Policy Optimization (GRPO) has become the core part of training Reasoning LLMs. However, we find some deficiency that influences RL stability and inference efficiency, like z…

Graph-GRPO: Stabilizing Multi-Agent Topology Learning via Group Relative Policy Optimization

2026-03-03 · Yueyang Cang, Xiaoteng Zhang, Erlu Zhao, Zehua Ji 외 arxiv

Optimizing communication topology is fundamental to the efficiency and effectiveness of Large Language Model (LLM)-based Multi-Agent Systems (MAS). While recent approaches utilize reinforcement learning to dynamically co…

Reinforcement LearningCode Generation

Hierarchy-of-Groups Policy Optimization for Long-Horizon Agentic Tasks

2026-02-26 · Shuo He, Lang Feng, Qi Wei, Xin Cheng 외 arxiv

Group-based reinforcement learning (RL), such as GRPO, has advanced the capabilities of large language models on long-horizon agentic tasks. To enable more fine-grained policy updates, recent research has increasingly sh…

Reinforcement Learning