paper-with-me

홈 › Papers

Reward Evolution with Graph-of-Thoughts: A Bi-Level Language Model Framework for Reinforcement Learning

2025-09-19 · Changwei Yao, Xinzi Liu, Chen Li, Marios Savvides arxiv

Designing effective reward functions remains a major challenge in reinforcement learning (RL), often requiring considerable human expertise and iterative refinement. Recent advances leverage Large Language Models (LLMs) for automated reward design, but these approaches are limited by hallucinations, reliance on human feedback, and challenges with handling complex, multi-step tasks. In this work, we introduce Reward Evolution with Graph-of-Thoughts (RE-GoT), a novel bi-level framework that enhances LLMs with structured graph-based reasoning and integrates Visual Language Models (VLMs) for automated rollout evaluation. RE-GoT first decomposes tasks into text-attributed graphs, enabling comprehensive analysis and reward function generation, and then iteratively refines rewards using visual feedback from VLMs without human intervention. Extensive experiments on 10 RoboGen and 4 ManiSkill2 tasks demonstrate that RE-GoT consistently outperforms existing LLM-based baselines. On RoboGen, our method improves average task success rates by 32.25%, with notable gains on complex multi-step tasks. On ManiSkill2, RE-GoT achieves an average success rate of 93.73% across four diverse manipulation tasks, significantly surpassing prior LLM-based approaches and even exceeding expert-designed rewards. Our results indicate that combining LLMs and VLMs with graph-of-thoughts reasoning provides a scalable and effective solution for autonomous reward evolution in RL.

📄 PDF Abstract BibTeX arXiv:2509.16136

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

MetaScale: Test-Time Scaling with Evolving Meta-Thoughts

2025-03-17 · Qin Liu, Wenxuan Zhou, Nan Xu, James Y. Huang 외

One critical challenge for large language models (LLMs) for making complex reasoning is their reliance on matching reasoning patterns from training data, instead of proactively selecting the most appropriate cognitive st…

Why Can Large Language Models Generate Correct Chain-of-Thoughts?

2023-10-20 · Rasul Tutunov, Antoine Grosnit, Juliusz Ziomek, Jun Wang 외

This paper delves into the capabilities of large language models (LLMs), specifically focusing on advancing the theoretical comprehension of chain-of-thought prompting. We investigate how LLMs can be effectively induced …

Text Generation

RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement Learning

2025-04-24 · Zihan Wang, Kangrui Wang, Qineng Wang, Pingyue Zhang 외

Training large language models (LLMs) as interactive agents presents unique challenges including long-horizon decision making and interacting with stochastic environment feedback. While reinforcement learning (RL) has en…

Decision MakingReinforcement Learning (RL)

Latent Thinking Optimization: Your Latent Reasoning Language Model Secretly Encodes Reward Signals in Its Latent Thoughts

2025-09-30 · Hanwen Du, Yuxin Dong, Xia Ning arxiv

Large Language Models (LLMs) excel at problem solving by generating chain of thoughts in natural language, but such verbal thinking is computationally costly and prone to overthinking. A recent work instead proposes a la…

Reinforced Graph of Thoughts: RL-Driven Adaptive Prompting for LLMs

2026-05-21 · Manuel Noah Riesen, Peter Alfred von Niederhäusern arxiv

Graph of Thoughts (GoT), a generalized form of recent prompting paradigms for large language models (LLMs), has been shown to be useful for elaborate problem solving. By executing a graph of operations, thoughts of the L…

Reinforcement Learning