**Chain-of-Thought Faithfulness of Reasoning Models Varies with Where and How Pr

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

> 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 mess...

**来源信息**
- **来源**：HuggingFace Daily Papers（社区热门论文）
- **分类**：论文
- **发布时间**：2026-09-01 08:00（北京时间）
- **原文**：[打开原文](https://huggingface.co/papers/2608.29464)