LLMs人格的机械论分析：通过潜在特征干预引导人格
Mechanistic Personality Analysis of LLMs: Steering Personality via Latent Feature Interventions
AI总结：
提出一种机械可解释性方法，通过稀疏自编码器和对比激活分析识别残差流中的潜在方向，并施加加法干预向量来增强目标OCEAN人格特质，同时保持语言建模性能。
摘要：
大型语言模型（LLMs）已展现出在生成文本中模拟类似人类OCEAN人格特质的能力。以往的工作侧重于通过提示工程或微调来塑造LLM人格。在这项工作中，我们提出了一种机械可解释性方法，直接干预模型的潜在特征。我们的方法使用稀疏自编码器（SAEs）和对比激活分析，识别残差流中对应于目标OCEAN特质的潜在方向。我们形式化了激活空间中的加法引导向量，并演示了如何对隐藏状态施加小的加法偏移来增强目标特质，同时保持整体语言建模性能。为了确定特征偏移的最优组合，我们探索了一种线性加权启发式方法，结合网格搜索优化，以平衡人格表达与任务性能。我们的方法在机械层面可控地引导人格特质，同时在标准基准上保持高性能，显示出潜力。
英文摘要：
Large Language Models (LLMs) have demonstrated the ability to simulate human-like OCEAN personality traits in generated text. Previous efforts have focused on prompt engineering or fine-tuning to shape LLM personality. In this work, we propose a mechanistic interpretability approach that directly intervenes on the model's latent features. Our method identifies latent directions in the residual stream corresponding to a target OCEAN trait using sparse autoencoders (SAEs) and contrastive activation analysis. We formalize an additive steering vector in activation space and demonstrate how applying a small additive shift to the hidden states enhances the target trait while preserving overall language modeling performance. To determine the optimal combination of feature shifts, we explore a linear weighting heuristic with grid search optimization that balances personality expression with task performance. Our approach shows promise in controllably steering personality traits at the mechanistic level while maintaining high performance on standard benchmarks.
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。
