paper-with-me

홈 › Papers

Unlocking Reasoning Capabilities in LLMs via Reinforcement Learning Exploration

2025-10-04 · Wenhao Deng, Long Wei, Chenglei Yu, Tailin Wu arxiv

Reinforcement learning with verifiable rewards (RLVR) has recently enhanced the reasoning capabilities of large language models (LLMs), particularly for mathematical problem solving. However, a fundamental limitation remains: as the sampling budget increases, the advantage of RLVR-trained models over their pretrained bases often diminishes or even vanishes, revealing a strong dependence on the base model's restricted search space. We attribute this phenomenon to the widespread use of the reverse Kullback-Leibler (KL) divergence regularizer, whose mode-seeking behavior keeps the policy trapped inside the base model's support region and hampers wider exploration. To address this issue, we propose RAPO (Rewards-Aware Policy Optimization), an algorithm to promote broader yet focused exploration. Our method (i) utilizes the forward KL penalty to replace the reverse KL penalty for out-of-distribution exploration, and (ii) reweights the reference policy to facilitate adaptive in-distribution exploration. We train Qwen2.5-3B and 7B models with RAPO on the 8K SimpleRL-Zero dataset, without supervised fine-tuning, and evaluate them on AIME2024 and AIME2025. Results show that RAPO consistently improves problem-solving performance. Notably, RAPO enables models to surpass the base model's performance ceiling and solves previously intractable problems, advancing the frontier of RLVR for challenging reasoning tasks.

📄 PDF Abstract BibTeX arXiv:2510.03865

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

Efficient and Stable Reinforcement Learning for Diffusion Language Models

2026-02-09 · Jiawei Liu, Xiting Wang, Yuanyuan Zhong, Defu Lian 외 arxiv

Reinforcement Learning (RL) is crucial for unlocking the complex reasoning capabilities of Diffusion-based Large Language Models (dLLMs). However, applying RL to dLLMs faces unique challenges in efficiency and stability.…

Reinforcement Learning

MARGE: Improving Math Reasoning for LLMs with Guided Exploration

2025-05-18 · Jingyue Gao, Runji Lin, Keming Lu, Bowen Yu 외

Large Language Models (LLMs) exhibit strong potential in mathematical reasoning, yet their effectiveness is often limited by a shortage of high-quality queries. This limitation necessitates scaling up computational respo…

MathMathematical Reasoning

Unlocking Exploration in RLVR: Uncertainty-aware Advantage Shaping for Deeper Reasoning

2025-10-12 · Can Xie, Ruotong Pan, Xiangyu Wu, Yunfei Zhang 외 arxiv

Reinforcement Learning with Verifiable Rewards (RLVR) has shown significant promise for enhancing the reasoning capabilities of large language models (LLMs). However, prevailing algorithms like GRPO broadcast a uniform a…

Reinforcement LearningMathematical Reasoning

Unlocking Aha Moments via Reinforcement Learning: Advancing Collaborative Visual Comprehension and Generation

2025-06-02 · Kaihang Pan, Yang Wu, Wendong Bu, Kai Shen 외

Recent endeavors in Multimodal Large Language Models (MLLMs) aim to unify visual comprehension and generation. However, these two capabilities remain largely independent, as if they are two separate functions encapsulate…

Image GenerationText to Image GenerationText-to-Image Generation

Tagging the Thought: Unlocking Personalization Reasoning via Reinforcement Learning

2025-09-27 · Song Jin, Juntian Zhang, Yong Liu, Xun Zhang 외 arxiv

Recent advancements have endowed Large Language Models (LLMs) with impressive general reasoning capabilities, yet they often struggle with personalization reasoning - the crucial ability to analyze user history, infer un…

Reinforcement Learning