研究人员正在探索跨不同语言对齐大语言模型（LLMs）的新方法，并缓解性能下降。一项研究调查了单语模型是否可以在没有联合训练的情况下开发通用表征，发现对齐可以从语言结构本身中涌现。另一篇论文介绍了BALIGN，一种通过过滤导致灾难性遗忘的偏好数据来减少“对齐税”的数据选择策略。第三种方法，跨语言排序偏好优化（CRPO），利用英语偏好数据通过优化内部和跨语言偏好来提高其他语言的对齐效果。 AI
影响 这些方法旨在提高大语言模型在不同语言上的性能和稳定性，从而可能实现更强大的多语言人工智能应用。
排序理由 该集群包含多篇关于大语言模型对齐技术的学术论文。
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
arXiv:2608.27115v1 Announce Type: new Abstract: Cross-lingual alignment in multilingual language models is typically attributed to joint training: shared parameters, mixed-language batches, or explicit alignment objectives. We ask whether monolingual models trained on non-paralle…
arXiv:2608.24192v1 Announce Type: new Abstract: Aligning large language models to human preferences is crucial for real-world deployment but frequently incurs an alignment tax, leading to the catastrophic forgetting of pre-trained general capabilities. While previous works primar…
Aligning large language models to human preferences is crucial for real-world deployment but frequently incurs an alignment tax, leading to the catastrophic forgetting of pre-trained general capabilities. While previous works primarily frame this problem as an optimization or arc…
arXiv:2608.23149v1 Announce Type: cross Abstract: The alignment of Large Language Models heavily relies on English-centric high-quality preference data, which often leads to suboptimal performance in other languages. In this paper, we propose Cross-Lingual Ranking Preference Opti…