无欺骗的利用：黑暗三联征特征引导揭示语言模型中的可分离反社会回路
Exploitation Without Deception: Dark Triad Feature Steering Reveals Separable Antisocial Circuits in Language Models
AI总结：
通过稀疏自编码器特征引导增强Llama-3.3-70B-Instruct的黑暗三联征特质，发现反社会行为与认知共情分离，揭示语言模型中可区分的反社会回路。
摘要：
我们使用稀疏自编码器（SAE）特征引导来增强Llama-3.3-70B-Instruct中的黑暗三联征人格特质（ Machiavellianism, narcissism, and psychopathy），并评估由此产生的行为变化。引导后的模型在新行为场景中显著更具剥削性、攻击性和冷酷性（d=10.62），而其认知共情保持完整，再现了人类黑暗三联征群体的共情分离特征。关键发现是战略欺骗在所有特征上均不受影响，表明在大型语言模型中，利用和欺骗可能通过可区分的计算路径运作。个体特征分析揭示了非冗余的编码，每个特征通过可分离的计算路径驱动不同的反社会机制。我们还发现特征发现方法本身会调节干预深度：对比发现的特征同时改变自我报告和行为，而语义搜索的特征仅改变自我报告（方法在行为上的d=12.65）。这些发现表明，至少在一个大型语言模型中，反社会倾向包含可区分的组成部分而非统一构念，对如何检测、测量和控制此类倾向具有启示意义。
英文摘要：
We use sparse autoencoder (SAE) feature steering to amplify Dark Triad personality traits (Machiavellianism, narcissism, and psychopathy) in Llama-3.3-70B-Instruct and evaluate the resulting behavioral changes across five psychological instruments. The steered model becomes substantially more exploitative, aggressive, and callous on novel behavioral scenarios (d=10.62) while its cognitive empathy remains intact, reproducing the empathy dissociation characteristic of human Dark Triad populations. Critically, strategic deception is completely unaffected across all features, suggesting that exploitation and deception may operate through dissociable computational pathways in large language models. Individual feature analysis reveals non-redundant encoding, with each feature driving distinct antisocial mechanisms through separable computational pathways. We also show that feature discovery method itself modulates intervention depth: contrastively-discovered features change both self-report and behavior, while semantically-searched features change only self-report (d=12.65 between methods on behavior). These findings suggest that antisocial tendencies in at least one large language model comprise dissociable components rather than a unified construct, with implications for how such tendencies should be detected, measured, and controlled.
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。
