用翻译模型扩展大模型多语言能力，不重训核心模型
Language on Demand, Knowledge at Core: Composing LLMs with Encoder-Decoder Translation Models for Extensible Multilinguality
- 将翻译模型作外挂，让大模型专注英语知识处理
- 在低资源语言上性能提升显著，最高达42.3%绝对增益
- 适合需要低成本扩展多语言的应用场景
大语言模型虽具强大通用智能，但多语言表现严重失衡。尽管其在统一语义空间中编码了大量跨语言知识，却常难以可靠地与低资源或未见语言交互。预训练的编码器-解码器翻译模型具备均衡的多语言能力，可自然补足大模型短板。本文提出XBridge：一种组合式编码器-大模型-解码器架构，将多语言理解与生成任务交由外部预训练翻译模型承担，同时保留大模型作为英语中心的知识处理核心。为解决模型间表征错位问题，引入轻量级跨模型映射层及基于最优传输的对齐目标，实现细粒度语义一致性。在四个大模型上开展的多语言理解、推理、摘要与生成实验表明，XBridge优于强基线，尤其在低资源及此前未见语言上表现突出，且无需重训练大模型。
原文摘要 · Abstract (English)
Large language models (LLMs) exhibit strong general intelligence, yet their multilingual performance remains highly imbalanced. Although LLMs encode substantial cross-lingual knowledge in a unified semantic space, they often struggle to reliably interface this knowledge with low-resource or unseen languages. Fortunately, pretrained encoder-decoder translation models already possess balanced multilingual capability, suggesting a natural complement to LLMs. In this work, we propose XBridge, a compositional encoder-LLM-decoder architecture that offloads multilingual understanding and generation to external pretrained translation models, while preserving the LLM as an English-centric core for general knowledge processing. To address the resulting representation misalignment across models, we introduce lightweight cross-model mapping layers and an optimal transport-based alignment objective, enabling fine-grained semantic consistency for multilingual generation. Experiments on four LLMs across multilingual understanding, reasoning, summarization, and generation indicate that XBridge outperforms strong baselines, especially on low-resource and previously unseen languages, without retraining the LLM.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。