ExpLang: 通过在线策略思考语言选择改进大语言模型推理中的探索与利用
ExpLang: Improved Exploration and Exploitation in LLM Reasoning with On-Policy Thinking Language Selection
AI总结:
ExpLang通过在线策略思考语言选择改进大语言模型推理中的探索与利用,实现多语言优势提升
摘要:
当前大型推理模型(LRMs)在经过强化学习(RL)后的训练后,在具有挑战性的任务上表现出强大的能力。然而,先前的工作主要集中在英语推理上,以期望获得最强的性能,尽管已经展示了多语言思考的潜在优势,以及全球用户对原生思考痕迹的需求。在本文中,我们提出了ExpLang,一种新颖的LLM后训练流水线,通过使用多种语言实现在线策略思考语言选择,以在RL过程中改进探索与利用。结果表明,我们的方法在相同的训练预算下稳定地优于仅使用英语训练,同时在已见过和未见过的语言中表现出高思考语言合规性。分析显示,通过在RL过程中将在线策略思考语言选择作为动作,ExpLang有效地扩展了RL探索空间,利用多样化语言偏好,并通过利用非英语优势提高了RL利用结果。该方法与大多数RL算法正交,并为利用多语言性来改进LRMs开辟了新的视角。
英文摘要:
Current large reasoning models (LRMs) have shown strong ability on challenging tasks after reinforcement learning (RL) based post-training. However, previous work mainly focuses on English reasoning in expectation of the strongest performance, despite the demonstrated potential advantage of multilingual thinking, as well as the requirement for native thinking traces by global users. In this paper, we propose ExpLang, a novel LLM post-training pipeline that enables on-policy thinking language selection to improve exploration and exploitation during RL with the use of multiple languages. The results show that our method steadily outperforms English-only training with the same training budget, while showing high thinking language compliance for both seen and unseen languages. Analysis shows that, by enabling on-policy thinking language selection as an action during RL, ExpLang effectively extends the RL exploration space with diversified language preference and improves the RL exploitation outcome with leveraged non-English advantage. The method is orthogonal to most RL algorithms and opens up a new perspective on using multilinguality to improve LRMs.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。