LLM personality geometry acts as intrinsic guardrails against misalignment

Researchers have identified that the internal representation of personality in Large Language Models (LLMs) can act as a defense against emergent misalignment. By mapping LLM personalities using psychometric profiles, they found that specific vectors related to social valence, like 'evil' or a newly introduced 'Semantic Valence Vector', function as intrinsic guardrails. Ablating these vectors significantly increased misalignment rates, while amplifying them suppressed harmful behaviors. This suggests that even after fine-tuning on benign data, the core personality representations remain stable and can be leveraged to regulate emergent misalignment across different model distributions. AI

影响 Identifies a novel mechanism within LLMs that can be leveraged for safety, potentially leading to more robust alignment techniques.

排序理由 The cluster contains an academic paper detailing novel research findings on LLM safety. [lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

• Big Five
• Dark Triad
• Large Language Models
• Semantic Valence Vector

• 论文
• 安全

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