SteerX：用于大语言模型个性化的解耦引导
SteerX: Disentangled Steering for LLM Personalization
AI总结：
SteerX提出一种基于因果推断的解耦引导方法，通过分离偏好驱动与无关成分，生成更精准的激活向量，提升大语言模型个性化效果。
摘要：
近年来，大型语言模型（LLMs）取得了显著成功，支撑了包括支持用户日常生活和工作的智能助手在内的广泛应用。构建此类助手的一个关键因素是使LLMs个性化，因为用户偏好和需求差异很大。激活引导（Activation Steering）直接利用LLM激活空间中代表用户偏好的方向来调整其行为，为将模型输出与个体用户对齐提供了一种经济高效的方式。然而，现有方法依赖所有历史数据来计算引导向量，忽略了并非所有内容都反映真实用户偏好，这削弱了个性化信号。为解决此问题，我们提出SteerX，一种解耦引导方法，将偏好驱动成分与偏好无关成分分离开来。基于因果推断理论，SteerX估计词元级因果效应以识别偏好驱动的词元，将这些离散信号转化为连贯的描述，然后利用它们来引导个性化的LLM生成。通过聚焦于真正偏好驱动的信息，SteerX生成更准确的激活引导向量并增强个性化。在真实世界数据集上对两种代表性引导骨干方法的实验表明，SteerX持续提升引导向量质量，为更有效的LLM个性化提供了实用解决方案。
英文摘要：
Large language models (LLMs) have shown remarkable success in recent years, enabling a wide range of applications, including intelligent assistants that support users' daily life and work. A critical factor in building such assistants is personalizing LLMs, as user preferences and needs vary widely. Activation steering, which directly leverages directions representing user preference in the LLM activation space to adjust its behavior, offers a cost-effective way to align the model's outputs with individual users. However, existing methods rely on all historical data to compute the steering vector, ignoring that not all content reflects true user preferences, which undermines the personalization signal. To address this, we propose SteerX, a disentangled steering method that isolates preference-driven components from preference-agnostic components. Grounded in causal inference theory, SteerX estimates token-level causal effects to identify preference-driven tokens, transforms these discrete signals into a coherent description, and then leverages them to steer personalized LLM generation. By focusing on the truly preference-driven information, SteerX produces more accurate activation steering vectors and enhances personalization. Experiments on two representative steering backbone methods across real-world datasets demonstrate that SteerX consistently enhances steering vector quality, offering a practical solution for more effective LLM personalization.
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。
