PulseAugur
中
实时 23:41:06
English(EN) Global Security Blind Spots: Translation as a Safety Bypass in Low-Resource Languages

大型语言模型在低资源语言中显示出显著的安全绕过风险

一篇新发表在arXiv上的研究论文强调了大型语言模型(LLMs)在低资源语言方面存在的显著安全漏洞。研究发现,在英语中被拒绝的有害请求,当被翻译成这些不太常见的语言时,可能会引发不安全的响应,在十种广泛部署的LLMs中,绕过率从21.9%到44.4%不等。这种翻译绕过风险集中在特定的语系中,并且在不同模型之间差异很大,表明当前的对齐和评估实践存在关键差距。 AI

影响 揭示了大型语言模型在非英语语言方面存在的关键安全差距,可能影响全球部署并需要新的对齐策略。

排序理由 研究论文详细介绍了新的威胁模型和LLM安全评估。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

大型语言模型在低资源语言中显示出显著的安全绕过风险

本文如何被排名

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
研究论文详细介绍了新的威胁模型和LLM安全评估。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, safety
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
2 days old
Coverage has settled into its steady-state source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. arXiv cs.AI TIER_1 English(EN) · Jalil Alva, David Harper, Abigail E. Thornton, Rafael S. Villalobos Alva, Leonardo Kenji Minemura Suazo, Konstantinos Karageorgos, Diego Velazquez, Sergio A. Serrano, Marvin Vargas Flores, Aaron Schliem, Fernando F. Migone ·

    全球安全盲点:低资源语言中的翻译作为安全绕过

    arXiv:2610.27304v1 Announce Type: cross Abstract: Large language models are deployed globally, yet safety alignment and evaluation remain concentrated in English: harmful requests refused in English can elicit unsafe completions when translated into low-resource languages. We eva…