多语言思考，而非更难：一种数据高效的框架，用于教授推理模型进行语言混合
Think Multilingual, Not Harder: A Framework for Analyzing and Teaching Code-Switched Reasoning
AI总结：
本文提出一种数据高效的框架，用于识别并教授推理模型更有效的语言混合行为，通过分析不同模型的语言混合行为，开发出有效的微调干预措施，提升推理模型的语言混合能力。
摘要：
近期在推理能力上的进展使大语言模型能够解决越来越复杂的数学、符号和逻辑任务。有趣的是，尽管推理模型通常训练生成单语文本，这些模型也观察到会进行语言混合（即混合语言）。先前的工作要么将语言混合视为不可取的错误，要么尝试通过修改输入提示或输出解码过程来控制语言混合，或专注于狭窄的语言、领域、任务和模型子集。我们通过引入第一个语言和行为动机的微调框架来填补这些空白，该框架用于识别大语言模型中的有益语言混合行为，并教授这些模型更有效地进行语言混合以进行推理。首先，我们创建并系统分析了来自不同模型、语言、任务和领域的推理轨迹数据集，以了解现有推理模型中发现的语言混合行为类型。然后，我们开发了微调干预措施，使推理模型能够根据现有模型中观察到的有益行为进行语言混合。我们发现，我们的框架可以以数据高效的方式显著增加有益的语言混合推理行为。有趣的是，我们还发现，推理模型中的语言混合行为可以通过微调任务（例如机器翻译）来修改。我们的工作表明，数据高效的干预可以给推理模型带来有益的语言混合行为。
英文摘要：
Recent developments in reasoning capabilities have enabled large language models to solve increasingly complex mathematical, symbolic, and logical tasks. Interestingly, while reasoning models are often trained to generate monolingual text, these models have also been observed to code-switch (i.e., mix languages). Prior works have either viewed code-switching as an undesirable error, attempted to control code-switching through modifications to input prompts or the output decoding process, or focus on narrow subsets of languages, domains, tasks, and models. We address these gaps by introducing the first linguistically and behaviorally motivated fine-tuning framework for identifying beneficial code-switched reasoning behaviors in large language models and teaching these models to code-switch more effectively for reasoning. We create the Code-Switched Reasoning (CoRe) corpus, consisting of (1) 7k reasoning traces from 15 models, 18 languages, 10 scripts, and diverse reasoning domains, providing insights into potentially helpful code-switching behaviors, and (2) 40 carefully curated datasets for training and evaluating six interventions for improving code-switching in reasoning across three models and seven languages, totaling 120 fine-tuning conditions. Across 80k+ reasoning traces from both language/culture-agnostic and -specific evaluations, English-dominated reasoning, semantically accurate code-switching, and a more even mix of languages are positively associated with correct answers, whereas dense switching is associated with incorrect answers. Moreover, we are able to elicit positive behaviors through fine-tuning tasks that do not directly demonstrate code-switching. Our work suggests that small but well-curated datasets can change how reasoning models code-switch, allowing us to reap the benefits of reasoning in the many languages that lack large-scale reasoning data.
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。
