安全、治理与可靠性 颠覆级 有讲解视频
发表时间
2026-09-02
arXiv
2609.00051

收录解读

现有安全对齐能诱导拒答,但有害输入如何被检测、传递并转换成拒绝行为仍不透明,也容易在越狱提示下失效。

论文识别出由有害检测头、安全神经元和拒绝头组成的多阶段安全电路,并用头部消融、神经元干预和激活补丁验证各阶段的因果关系,再以结构不变的权重缩放增强电路。

跨Llama、Mistral、Guanaco和Qwen等六个模型、GCG及多种附加攻击评估后,平均安全率提升26.5%,四项能力基准平均仅下降1.7%;盲评和组件稳定性实验进一步支持结果。

工作同时给出可解释机制和低成本干预方式,在安全工程与机械可解释性之间形成可迁移桥梁;但电路是静态快照,模型更新和自适应攻击下的长期稳定性尚未知。

原始摘要

Despite extensive alignment efforts, Large Language Models (LLMs) remain vulnerable to generating unsafe content under adversarial prompting, yet the internal mechanisms by which safety behaviors are implemented remain poorly understood. We study LLM safety from a mechanistic interpretability perspective and characterize a multi-stage safety circuit that organizes refusal behavior, consisting of (i) Harmful Detection Heads that respond to harmful inputs, (ii) Safety Neurons that mediate and stabilize safety signals in the residual stream, and (iii) Refusal Heads that translate these signals into safe response generation. Using targeted attention-head and neuron-level interventions, we provide causal evidence consistent with this circuit organization, showing that suppressing upstream Harmful Detection Heads disrupts downstream refusal behavior and that safety neurons mediate this interaction. We validate that this decomposition recurs across multiple LLM architectures and adversarial attack settings, and use simple, architecture-preserving weight scaling as a mechanistic probe to test its functional relevance. Across six LLMs, circuit-guided scaling improves safety rates under attacks by 26.5%, while incurring only a 1.7% accuracy drop across four standard benchmarks. Overall, our results support a circuit-level interpretation of LLM safety and suggest that mechanistic abstractions can reveal stable and transferable patterns underlying aligned behavior. Code is available at https: //github.com/Trustworthy-ML-Lab/ Detection2Refusal.

解读视频

相关论文

链接