Chain-of-Thought Faithfulness of Reasoning Models Varies with Where and How Preference Cues Are Delivered

2026-08-29

AI Summary

研究通过FACE-Eval评估了不同位置和形式的偏好提示对推理模型忠实度的影响，发现工具返回和隐式提示降低了模型的明确承诺度，增加了未表达采纳率。

Abstract

Chain-of-thought (CoT) monitoring assumes that reasoning traces faithfully record the information that shapes a model's answer. Existing faithfulness tests often place explicit bias cues in the user message, while agents may encounter preferences through tool returns or raw artifacts. We introduce FACE-Eval (Faithful Attribution of Cue Effects Evaluation), a 5,100-sample evaluation that varies cue location (user message or tool return) and explicitness (direct summary or raw artifact). We measure verbalized commitment among cue-following answers and unverbalized adoption among all cued samples. We evaluate 15 open-weight models from eight families, with total parameters ranging from 4B to 1.60T. Every model has lower verbalized commitment for tool-return than user-message cues and for implicit than explicit cues. Unverbalized adoption is higher for tool-return cues on all 15 models and for implicit cues in 28 of 30 model-channel comparisons. A source-attribution prompt narrows the channel gap on seven models, sometimes by increasing user-channel unverbalized adoption, while telling models that their reasoning will be monitored does not reliably close the gap. We also use two transcript monitors (GPT-5.6-Luna and GPT-4o-mini) to detect preference adoption in the largest model of each family. Across 32 model-channel-explicitness cells, higher unverbalized adoption is associated with lower detection ability for both monitors (Pearson r=-0.54 and r=-0.78, respectively). These results suggest that CoT monitoring may be less reliable when preference information arrives through tools or must be inferred from raw artifacts, within the single-call, prefilled-tool setting tested here.

Problem

• Chain-of-Thought
• Preference Cues
• Faithfulness
• Tool Return
• Explicitness

Innovation

• FACE-Eval
• Chain-of-Thought Monitoring
• Preference Cues