用激活控制让AI更诚实说出推理关键步骤，效果跨数据集通用。
On the Generalization of Steering Vectors for Chain-of-Thought Faithfulness
- 通过激活向量引导模型承认提示中的关键线索。
- 大模型（Gemma-3 12B）在多种提示下承认率提升，效果稳定。
- 方法构建方式不影响效果，且不改变使用频率，只让隐藏使用变透明。
随着链式思维（CoT）规模扩大，模型能力显著提升，这为人工智能安全带来希望——通过让模型自述推理过程，可实现监控。然而，某些情况下模型会忽略推理中关键步骤。例如，当提示包含错误答案线索时，模型可能不承认该线索对其结论的影响。这种未披露关键推理步骤的现象称为不忠实。已有研究显示激活控制可提升链式思维的忠实性。本文扩展此方向，考察三种模型（Gemma-3 4B、Qwen-3.5 9B、Gemma-3 12B）在提示问答任务中，不同提示类型、数据集及向量构建方法下，激活控制对忠实性的泛化表现。结果显示，仅在最大模型Gemma-3 12B上，激活控制能可靠提升线索承认率；但一旦有效，其效果在跨提示类型和跨数据集测试中均表现一致，主要受评估设置影响，而非训练设置。四种向量构造方法（包括一种不明确指定提示的目标）产生的效果大小相近。进一步分析表明，激活控制并未提高线索使用频率，而是减少未被承认的线索使用，即‘隐藏使用’，说明其作用在于增强显式表达而非增加使用。结果支持激活控制作为提升模型推理透明性的有效手段。
原文摘要 · Abstract (English)
Model capabilities have improved in large part due to scaling chain of thought. This has been a promising development for AI safety--where models verbalize their reasoning, it is possible to monitor it. However, in some cases, models do not verbalize important steps in their reasoning process. For example, models prompted with a cue suggesting the incorrect answer may fail to acknowledge that cue, even when it appears instrumental to their conclusion. When chain of thought (CoT) fails to disclose instrumental reasoning steps, we describe it as unfaithful. Prior work has shown that activation steering can be a useful method to improve faithfulness in CoT. We extend this line of work by studying how well steering for faithfulness generalizes across cue types, datasets, and methods of constructing the steering vector for three models (Gemma-3 4B, Qwen-3.5 9B, Gemma-3 12B) in a cued question-answering setting. While steering reliably increases cue acknowledgment for only the largest model (Gemma-3 12B), we find that when steering is effective, its effect generalizes broadly across cue types and datasets--in cross-cue and cross-dataset analyses, effect size is determined primarily by the evaluation setting, rather than the vector's train setting. How the vector is built also matters little--four construction methods, including one whose optimization target mentions no specific cue, yield similar effect sizes. Finally, we consider the possibility that steering promotes the salience of the cue and causes greater cue use, rather than targeting verbalization behaviors. However, we find no evidence for this--steering leaves the rate of cue use roughly unchanged while reducing hidden cue use, i.e., cue use that is not acknowledged.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。