拒绝引导：为敏感话题对LLM拒绝行为实现细粒度控制
Refusal Steering: Fine-grained Control over LLM Refusal Behaviour for Sensitive Topics
AI总结：
通过引导向量控制LLM在敏感话题上的拒绝行为，实现安全且可控的审核方法。
摘要：
我们引入了拒绝引导，这是一种推理时的方法，用于在不重新训练的情况下对大型语言模型在政治敏感话题上的拒绝行为进行细粒度控制。我们用一个LLM作为判断者来替代脆弱的基于模式的拒绝检测，该判断者会分配拒绝置信度分数，并提出一个岭正则化的变体来计算引导向量，以更好地隔离拒绝-顺从方向。在Qwen3-Next-80B-A3B-Thinking上，我们的方法在政治敏感话题上消除了模型的拒绝行为，同时在JailbreakBench上保持了安全性和在通用基准上接近基线性能。该方法在4B和80B模型上具有泛化能力，也可以在需要时诱导定向拒绝。我们分析了引导向量并展示拒绝信号集中在transformer的深层层中，并分布在许多维度上。这些结果共同表明，激活引导可以消除政治拒绝行为，同时保留对有害内容的安全对齐，为推理时可控、透明的审核提供了一条实际路径。
英文摘要：
We introduce Refusal Steering, an inference-time method to exercise fine-grained control over Large Language Models refusal behaviour on politically sensitive topics without retraining. We replace fragile pattern-based refusal detection with an LLM-as-a-judge that assigns refusal confidence scores and we propose a ridge-regularized variant to compute steering vectors that better isolate the refusal--compliance direction. On Qwen3-Next-80B-A3B-Thinking, our method removes the refusal behaviour of the model around politically sensitive topics while maintaining safety on JailbreakBench and near-baseline performance on general benchmarks. The approach generalizes across 4B and 80B models and can also induce targeted refusals when desired. We analize the steering vectors and show that refusal signals concentrate in deeper layers of the transformer and are distributed across many dimensions. Together, these results demonstrate that activation steering can remove political refusal behaviour while retaining safety alignment for harmful content, offering a practical path to controllable, transparent moderation at inference time.
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。
