组合式多语言与行为属性引导
Compositional Multilingual and Behavioral Attribute Steering
浏览论文内容
中文总结 AI 辅助
本研究证明在大型语言模型中，将属性引导向量分别注入其最优层可实现无训练的组合式多属性控制，且向量在残差流中近似正交。
中文摘要 AI 辅助
本研究探讨了大型语言模型中用于语言和行为控制的引导向量的组合性。我们聚焦于语言、越狱和简洁性这三个属性，研究了无训练的加法式属性引导向量组合是否能在来自两个模型家族、两种规模尺度的四个指令微调模型中保持每个属性的预期引导效果。我们发现，单一属性引导对所有三个属性都是可靠的，但仅在适当的干预层和引导强度的组合下才有效，其中抽象行为（越狱、简洁性）偏好中间层，而语言偏好较早的层。我们表明，当两个属性向量各自注入其表现最佳的层时，这两个属性向量的加法组合能成功同时引导这两个属性，并且这种效果部分扩展到三个同时组合的属性，解决了先前关于无训练组合工作遗留的不一致性问题。我们进一步分析了这些引导向量的几何特性，发现它们在残差流中近似正交，这与其组合行为一致。
英文摘要
This study examines the compositionality of steering vectors for language and behavioral control in large language models. Focusing on language, jailbreak, and conciseness, we investigate whether additive, training-free composition of attribute steering vectors can preserve the intended steering effect of each attribute, across four instruction-tuned models from two model families and two size scales. We find that single-attribute steering is reliable for all three attributes, but only within an appropriate combination of intervention layer and steering strength, with abstract behaviors (jailbreak, conciseness) favoring middle layers and language favoring earlier layers. We show that additive composition of two attribute vectors succeeds in steering both attributes simultaneously when each is injected at its own best-performing layer, and that this partially extends to three simultaneously composed attributes, addressing an inconsistency left open by prior work on training-free composition. We further analyze the geometric properties of these steering vectors, finding that they are approximately orthogonal in the residual stream, consistent with their compositional behavior.
发表机构
- German Research Centre for Artificial Intelligence (DFKI)（德国人工智能研究中心）
- Saarland University（萨尔大学）
- Centre for European Research in Trusted AI (CERTAIN)（欧洲可信人工智能研究中心）
机构由 AI 辅助整理，请以论文原文为准。
