paper-with-me

Papers

Tool-Augmented Reward Modeling

2023-10-02 · Lei LI, Yekun Chai, Shuohuan Wang, Yu Sun, Hao Tian, Ningyu Zhang, Hua Wu

Reward modeling (a.k.a., preference modeling) is instrumental for aligning large language models with human preferences, particularly within the context of reinforcement learning from human feedback (RLHF). While conventional reward models (RMs) have exhibited remarkable scalability, they oft struggle with fundamental functionality such as arithmetic computation, code execution, and factual lookup. In this paper, we propose a tool-augmented preference modeling approach, named Themis, to address these limitations by empowering RMs with access to external environments, including calculators and search engines. This approach not only fosters synergy between tool utilization and reward grading but also enhances interpretive capacity and scoring reliability. Our study delves into the integration of external tools into RMs, enabling them to interact with diverse external sources and construct task-specific tool engagement and reasoning traces in an autoregressive manner. We validate our approach across a wide range of domains, incorporating seven distinct external tools. Our experimental results demonstrate a noteworthy overall improvement of 17.7% across eight tasks in preference ranking. Furthermore, our approach outperforms Gopher 280B by 7.3% on TruthfulQA task in zero-shot evaluation. In human evaluations, RLHF trained with Themis attains an average win rate of 32% when compared to baselines across four distinct tasks. Additionally, we provide a comprehensive collection of tool-related RM datasets, incorporating data from seven distinct tool APIs, totaling 15,000 instances. We have made the code, data, and model checkpoints publicly available to facilitate and inspire further research advancements\footnote{\url{https://github.com/ernie-research/Tool-Augmented-Reward-Model}}.

📄 PDF Abstract BibTeX arXiv:2310.01045

Code (1)

ernie-research/Tool-Augmented-Reward-Model 공식 구현 pytorch

Tasks

TruthfulQA

Similar Papers 제목 키워드 기반

OpenReward: Learning to Reward Long-form Agentic Tasks via Reinforcement Learning

2025-10-28 · Ziyou Hu, Zhengliang Shi, Minghang Zhu, Haitao Li 외 arxiv

Reward models (RMs) have become essential for aligning large language models (LLMs), serving as scalable proxies for human evaluation in both training and inference. However, existing RMs struggle on knowledge-intensive …

Reinforcement Learning

AgentV-RL: Scaling Reward Modeling with Agentic Verifier

2026-04-17 · Jiazheng Zhang, Ziche Fu, Zhiheng Xi, Wenqing Jing 외 arxiv

Verifiers have been demonstrated to enhance LLM reasoning via test-time scaling (TTS). Yet, they face significant challenges in complex domains. Error propagation from incorrect intermediate reasoning can lead to false p…

Reinforcement Learning

Tool-Augmented Agent for Closed-loop Optimization,Simulation,and Modeling Orchestration

2026-04-01 · Liyuan Deng, Shujian Deng, Yongkang Chen, Yongkang Dai 외 arxiv

Iterative industrial design-simulation optimization is bottlenecked by the CAD-CAE semantic gap: translating simulation feedback into valid geometric edits under diverse, coupled constraints. To fill this gap, we propose…

Reinforcement Learning

COSMO-Agent: Tool-Augmented Agent for Closed-loop Optimization,Simulation,and Modeling Orchestration

2026-04-07 · Liyuan Deng, Shujian Deng, Yongkang Chen, Yongkang Dai 외 arxiv

Iterative industrial design-simulation optimization is bottlenecked by the CAD-CAE semantic gap: translating simulation feedback into valid geometric edits under diverse, coupled constraints. To fill this gap, we propose…

Reinforcement Learning

Rethinking Reinforcement Fine-Tuning in LVLM: Convergence, Reward Decomposition, and Generalization

2026-04-21 · Carter Adams, Rafael Oliveira, Gabriel Almeida, Sofia Torres arxiv

Reinforcement fine-tuning with verifiable rewards (RLVR) has emerged as a powerful paradigm for equipping large vision-language models (LVLMs) with agentic capabilities such as tool use and multi-step reasoning. Despite …