心理概念神经元：神经控制偏差探测与移位生成在大语言模型中的可能性？
Psychological Concept Neurons: Can Neural Control Bias Probing and Shift Generation in LLMs?
AI总结：
研究通过分析大五人格特质的内部表示，探讨其在大语言模型中的定位与行为输出的关系，发现通过干预特定神经元可影响模型的潜在表示和标签生成，但效果受限于概念依赖性和跨特质溢出。
摘要：
利用如大五人格特质等心理学构念，大语言模型（LLMs）可以模仿特定的人格特征并预测用户的人格。尽管LLMs可以表现出与这些构念一致的行为，但尚不清楚这些构念在模型内部如何表示以及如何影响行为输出。为解决这一差距，我们专注于问卷操作化的大五构念，分析其内部表示的形成和定位，并通过干预来检验这些表示如何与行为输出相关联。在我们的实验中，我们首先使用探测法来研究大五信息在模型深度中的出现位置。然后识别出对每个大五构念有选择性反应的神经元，并测试增强或抑制其激活是否能偏转潜在表示和标签生成。我们发现大五信息在早期层迅速可解码，并在最终层仍可检测到，而概念选择性神经元最常见于中层，并在不同领域间表现出有限的重叠。对这些神经元的干预一致地将探测结果转向目标构念，针对某些构念的成功率超过0.8，表明模型内部对大五人格特质的分离可以被因果地操控。在标签生成层面，相同的干预往往能偏转生成的标签分布，但效果较弱，更依赖于构念，并常伴随跨特质溢出，表明即使对大量概念选择性神经元进行干预，对生成标签的控制仍具有挑战性。总体而言，我们的发现揭示了大语言模型中表示控制与行为控制之间的差距。
英文摘要：
Using psychological constructs such as the Big Five, large language models (LLMs) can imitate specific personality profiles and predict a user's personality. While LLMs can exhibit behaviors consistent with these constructs, it remains unclear where and how they are represented inside the model and how they relate to behavioral outputs. To address this gap, we focus on questionnaire-operationalized Big Five concepts, analyze the formation and localization of their internal representations, and use interventions to examine how these representations relate to behavioral outputs. In our experiment, we first use probing to examine where Big Five information emerges across model depth. We then identify neurons that respond selectively to each Big Five concept and test whether enhancing or suppressing their activations can bias latent representations and label generation in intended directions. We find that Big Five information becomes rapidly decodable in early layers and remains detectable through the final layers, while concept-selective neurons are most prevalent in mid layers and exhibit limited overlap across domains. Interventions on these neurons consistently shift probe readouts toward targeted concepts, with targeted success rates exceeding 0.8 for some concepts, indicating that the model's internal separation of Big Five personality traits can be causally steered. At the label-generation level, the same interventions often bias generated label distributions in the intended directions, but the effects are weaker, more concept-dependent, and often accompanied by cross-trait spillover, indicating that comparable control over generated labels is difficult even with interventions on a large fraction of concept-selective neurons. Overall, our findings reveal a gap between representational control and behavioral control in LLMs.
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。
