Building Multilingual Bridges: Data Mixing as the Pillar of Generalization for In-Language Reasoning

  • 类型：arxiv
  • 标识：2609.10445
  • 链接：https://arxiv.org/abs/2609.10445
  • 主分类：llm-infra
  • 形态：method
  • 被引：0
  • 被引来源：Semantic Scholar + OpenAlex
  • S2被引：0
  • OpenAlex被引：0
  • 影响力被引：0
  • TLDR：It is shown the path to generalizing L2 reasoning to held-out languages goes through broader language coverage, readily available multilingual non-reasoning data, and a sufficient English reasoning backbone, indicating that reasoning is a language-agnostic behavior that can be transferred across typologically diverse languages through careful data mixing.
  • OpenAlex ID：W7212187918
  • OpenAlex DOI：10.48550/arxiv.2609.10445
  • DOI：10.48550/arxiv.2609.10445
  • DOI来源：OpenAlex
  • 开放获取：green
  • 开放获取链接：https://doi.org/10.48550/arxiv.2609.10445
  • OpenAlex更新：2026-09-13
  • 待LLM分类：否
  • 标题中文：构建多语言桥梁：数据混合作为语言内推理泛化的支柱
  • TLDR中文：表明将 L2 推理泛化到未见语言的关键路径在于更广泛的语言覆盖、现成可用的多语言非推理数据，以及足够强的英语推理骨干，表明推理是一种与语言无关的行为，可通过精心数据混合在类型多样的语言间迁移。
  • 成熟度：research
  • 场景：multilingual、data mixing、reasoning
  • 来源文件：
  • /inbox/tom/_candidates/2026-09-12-agent-rag-longcontext-candidates.json
  • [OpenAlex backfill]
  • /inbox/tom/_candidates/2026-09-13-agent-rag-longcontext-candidates.json
  • /inbox/tom/_candidates/2026-09-14-agent-rag-longcontext-candidates.json
  • /inbox/tom/_candidates/2026-09-14-agent-memory-tool-use-candidates.json
  • [S2 enrich]