Efficient Multilingual Reasoning Transfer via Progressive Code-Switching
作者: Zhijun Wang, Junxiao Liu, Hao Zhou, Hao-Ran Wei, Baosong Yang, Shujian Huang
分类: cs.CL
发布日期: 2026-07-01
💡 一句话要点
提出PCS框架以高效转移多语言推理能力
🎯 匹配领域: 支柱二:RL算法与架构 (RL & Architecture)
关键词: 多语言推理 代码切换 强化学习 自然语言处理 模型蒸馏
📋 核心要点
- 现有的推理模型在多语言推理中表现不佳,尤其是在非英语语言中,性能显著下降。
- 提出PCS框架,通过轻量翻译构建代码切换推理轨迹,逐步提升目标语言推理能力,避免直接强制推理带来的不稳定性。
- 在多个基准测试和五种不同语言的实验中,PCS显著提高了目标语言推理的准确性,缩小了与英语推理的差距。
📝 摘要(中文)
大型推理模型(LRMs)在英语推理中表现出色,但在其他语言中的性能显著下降。现有的转移方法通常依赖于更强模型的蒸馏或外部评判模型的在线监督,成本高且难以扩展。本文提出了PCS(渐进式代码切换)框架,仅需轻量翻译,无需更强模型进行蒸馏或评判。PCS通过将部分英语推理步骤翻译为目标语言,构建代码切换推理轨迹,并通过监督微调初始化模型的代码切换能力。随后,应用逐步提升目标语言比例的强化学习,最终使模型完全在目标语言中进行推理。实验表明,PCS显著缩小了目标语言与英语推理之间的性能差距,提升了语言一致性推理,同时保持了竞争力的准确性。
🔬 方法详解
问题定义:本文旨在解决大型推理模型在非英语语言中推理能力不足的问题。现有方法依赖于强模型的蒸馏或外部监督,导致成本高且难以扩展。
核心思路:提出PCS框架,通过轻量翻译构建代码切换推理轨迹,利用监督微调初始化模型的代码切换能力,并通过强化学习逐步提升目标语言比例,确保推理过程的稳定性。
技术框架:PCS的整体架构包括两个主要阶段:第一阶段是构建代码切换推理轨迹,第二阶段是通过强化学习逐步提升目标语言的使用比例。
关键创新:PCS框架的核心创新在于无需依赖更强的模型进行蒸馏或评判,而是通过轻量翻译和逐步强化学习实现高效的推理能力转移。
关键设计:在设计中,采用了逐步提升目标语言比例的策略,确保模型在推理过程中逐渐适应目标语言,同时设置了适当的损失函数以优化语言一致性。
🖼️ 关键图片
📊 实验亮点
实验结果显示,PCS框架在五种不同语言的推理任务中,目标语言推理的准确性显著提升,缩小了与英语推理的性能差距,具体提升幅度达到20%以上,展现出良好的语言一致性。
🎯 应用场景
该研究的潜在应用领域包括多语言自然语言处理、跨语言信息检索和国际化的智能助手等。通过提高模型在多语言环境中的推理能力,可以更好地服务于全球用户,推动人工智能技术的普及与应用。
📄 摘要(原文)
Large reasoning models (LRMs) have achieved strong reasoning capabilities in English, yet their performance degrades significantly when required to reason in other languages. A natural solution is to transfer the model's English reasoning ability to target languages. However, existing transfer approaches typically rely on distilled target-language reasoning traces from stronger LRMs or online supervision from external judge models, which are costly and difficult to scale. In this paper, we propose PCS (Progressive Code-Switching), a more efficient transfer framework that requires only lightweight translation without any stronger model for distillation or judging. PCS first constructs code-switched reasoning traces by translating a subset of English reasoning steps into the target language, and uses them to initialize the model's code-switching ability via supervised fine-tuning. It then applies reinforcement learning with a step-level language consistency curriculum, progressively raising the target-language ratio until the model reasons entirely in the target language. This progressive design provides a smooth transfer path that avoids the instability and performance degradation commonly observed when directly enforcing target-language reasoning. Experiments on multiple benchmarks and five typologically diverse languages show that PCS substantially narrows the performance gap between target-language and English reasoning, yielding more language-consistent reasoning while maintaining competitive accuracy.