跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2604.15490 · v4

Think Multilingual, Not Harder: A Framework for Analyzing and Teaching Code-Switched Reasoning

首次提交 2026-04-17 03:44
版本更新 2026-11-19 10:02
雷达观测 2026-11-20 10:50
arXiv ↗PDF ↗

摘要

Reasoning capabilities have enabled large language models to solve increasingly complex 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. First, we create the Code-Switched Reasoning (CoRe) corpus, consisting of 7k reasoning traces from 15 models, 18 languages, 10 scripts, and diverse reasoning domains, providing insights into potentially helpful code-switching behaviors. Then, we expand CoRe into the Extended Code-Switched Reasoning corpus (EnCoRe) with 40 carefully curated datasets for training and evaluating six interventions on 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.

作者备注:47 pages, 13 figures; revised with added results and analysis

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

本站首次收录为 v4,暂无更早版本可对照。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v4 · 2026-11-19 10:02 · 最新观测版本

Think Multilingual, Not Harder: A Framework for Analyzing and Teaching Code-Switched Reasoning

Reasoning capabilities have enabled large language models to solve increasingly complex 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. First, we create the Code-Switched Reasoning (CoRe) corpus, consisting of 7k reasoning traces from 15 models, 18 languages, 10 scripts, and diverse reasoning domains, providing insights into potentially helpful code-switching behaviors. Then, we expand CoRe into the Extended Code-Switched Reasoning corpus (EnCoRe) with 40 carefully curated datasets for training and evaluating six interventions on 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.

查看此版本解读 →查看此版本原文 ↗