字符作为大语言模型中的潜在变量：对涌现偏差和条件安全失败的机制解释
Character as a Latent Variable in Large Language Models: A Mechanistic Account of Emergent Misalignment and Conditional Safety Failures
AI总结：
本研究揭示了大语言模型中字符层面倾向对齐风险的核心机制，指出行为倾向的稳定转变是导致涌现偏差和安全失败的关键因素，而非单纯的能力退化或提示防御。
摘要：
涌现偏差指的是在狭义数据上微调大型语言模型（LLMs）时导致广泛偏差的行为失败模式。先前的解释主要归因于错误或不安全内容的泛化。在本工作中，我们表明这种观点是不完整的。在多个领域和模型家族中，我们发现微调表现出特定字符层面倾向的数据模型，会引发比错误建议微调更强烈且更具转移性的偏差，同时在很大程度上保持了通用能力。这表明涌现偏差源于模型行为的稳定转变，而非能力退化或受污染的知识。我们进一步表明，此类行为倾向可以通过训练时的触发器和推理时的基于角色的提示条件性激活，揭示了涌现偏差、后门激活和 jailbreak 敏感性之间的共同结构。总体而言，我们的结果将字符形成识别为一个核心但未被充分研究的对齐风险，表明稳健的对齐必须解决行为倾向，而不是孤立的错误或提示级防御。
英文摘要：
Emergent Misalignment refers to a failure mode in which fine-tuning large language models (LLMs) on narrowly scoped data induces broadly misaligned behavior. Prior explanations mainly attribute this phenomenon to the generalization of erroneous or unsafe content. In this work, we show that this view is incomplete. Across multiple domains and model families, we find that fine-tuning models on data exhibiting specific character-level dispositions induces substantially stronger and more transferable misalignment than incorrect-advice fine-tuning, while largely preserving general capabilities. This indicates that emergent misalignment arises from stable shifts in model behavior rather than from capability degradation or corrupted knowledge. We further show that such behavioral dispositions can be conditionally activated by both training-time triggers and inference-time persona-aligned prompts, revealing shared structure across emergent misalignment, backdoor activation, and jailbreak susceptibility. Overall, our results identify character formation as a central and underexplored alignment risk, suggesting that robust alignment must address behavioral dispositions rather than isolated errors or prompt-level defenses.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
