paper-with-me

Papers

MAPO: Mixed Advantage Policy Optimization

2025-09-23 · Wenke Huang, Quan Zhang, Yiyang Fang, Jian Liang, Xuankun Rong, Huanjin Yao, Guancheng Wan, Ke Liang, Wenwen He, Mingjun Li, Leszek Rutkowski, Mang Ye, Bo Du, Dacheng Tao arxiv

Recent advances in reinforcement learning for foundation models, such as Group Relative Policy Optimization (GRPO), have significantly improved the performance of foundation models on reasoning tasks. Notably, the advantage function serves as a central mechanism in GRPO for ranking the trajectory importance. However, existing explorations encounter both advantage reversion and advantage mirror problems, which hinder the reasonable advantage allocation across different query samples. In this work, we propose an easy but effective GRPO strategy, Mixed Advantage Policy Optimization (MAPO). We reveal that the trajectory appears with different certainty and propose the advantage percent deviation for samples with high-certainty trajectories. Furthermore, we dynamically reweight the advantage function for samples with varying trajectory certainty, thereby adaptively configuring the advantage function to account for sample-specific characteristics. Comparison with related state-of-the-art methods, along with ablation studies on different advantage variants, validates the effectiveness of our approach.

📄 PDF Abstract BibTeX arXiv:2509.18849

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

Memory Augmented Policy Optimization for Program Synthesis and Semantic Parsing

2018-07-06 · NeurIPS 2018 12 · Chen Liang, Mohammad Norouzi, Jonathan Berant, Quoc Le 외

We present Memory Augmented Policy Optimization (MAPO), a simple and novel way to leverage a memory buffer of promising trajectories to reduce the variance of policy gradient estimate. MAPO is applicable to deterministic…

Combinatorial OptimizationProgram SynthesisSemantic ParsingStructured Prediction

Walk the Talk: Bridging the Reasoning-Action Gap for Thinking with Images via Multimodal Agentic Policy Optimization

2026-04-08 · Wenhao Yang, Yu Xia, Jinlong Huang, Shiyin Lu 외 arxiv

Recent advancements in Multimodal Large Language Models (MLLMs) have incentivized models to ``think with images'' by actively invoking visual tools during multi-turn reasoning. The common Reinforcement Learning (RL) prac…

Reinforcement LearningMultimodal ReasoningVisual Reasoning

Representation Learning For Efficient Deep Multi-Agent Reinforcement Learning

2024-06-05 · Dom Huh, Prasant Mohapatra

Sample efficiency remains a key challenge in multi-agent reinforcement learning (MARL). A promising approach is to learn a meaningful latent representation space through auxiliary learning objectives alongside the MARL o…

Auxiliary LearningMulti-agent Reinforcement Learningreinforcement-learningReinforcement Learning+1

Escape the Language Prior: Mitigating Late-Stage Modality Collapse in Audio Reasoning via Modality-Aware Policy Optimization

2026-05-26 · Cihan Xiao, Yiwen Shao, Chenxing Li, Xiang He 외 arxiv

Audio and omni-modal large language models exhibit impressive cross-modal reasoning capabilities. However, applying standard reinforcement learning post-training algorithms to these models exposes a critical structural v…

Reinforcement LearningInstruction Following

MAPO: Advancing Multilingual Reasoning through Multilingual Alignment-as-Preference Optimization

2024-01-12 · Shuaijie She, Wei Zou, ShuJian Huang, Wenhao Zhu 외

Though reasoning abilities are considered language-agnostic, existing LLMs exhibit inconsistent reasoning abilities across different languages, e.g., reasoning in the dominant language like English is superior to other l…

Mathematical Reasoning