标题： 通过提示和稀疏自编码器引导LLM值的因果图导向调整 显示英文标题

标题： Causal Graph Guided Steering of LLM Values via Prompts and Sparse Autoencoders

摘要： 随着大型语言模型（LLMs）越来越多地融入关键应用，使其行为与人类价值观保持一致面临着重大挑战。 当前的方法，如基于人类反馈的强化学习（RLHF），通常只关注有限的一组价值观，并且资源消耗较大。 此外，价值观之间的相关性在很大程度上被忽视，尚未得到充分利用。 我们的框架通过挖掘一个因果图来解决这一局限性，该因果图阐明了LLMs中各种价值观之间的隐含关系。 利用因果图，我们实现了两种轻量级的价值引导机制：提示模板引导和稀疏自编码器特征引导，并分析了改变一个价值维度对其他维度的影响。 在Gemma-2B-IT和Llama3-8B-IT上进行的大量实验证明了我们引导方法的有效性和可控性。 显示英文摘要

摘要： As large language models (LLMs) become increasingly integrated into critical applications, aligning their behavior with human values presents significant challenges. Current methods, such as Reinforcement Learning from Human Feedback (RLHF), often focus on a limited set of values and can be resource-intensive. Furthermore, the correlation between values has been largely overlooked and remains underutilized. Our framework addresses this limitation by mining a causal graph that elucidates the implicit relationships among various values within the LLMs. Leveraging the causal graph, we implement two lightweight mechanisms for value steering: prompt template steering and Sparse Autoencoder feature steering, and analyze the effects of altering one value dimension on others. Extensive experiments conducted on Gemma-2B-IT and Llama3-8B-IT demonstrate the effectiveness and controllability of our steering methods.