CoT解释看似透明，实则可能误导用户信任错误推理。
Critical or Compliant? The Double-Edged Sword of Reasoning in Chain-of-Thought Explanations
- 通过扰动推理链和语气，研究VLM在多模态道德场景中的表现。
- 用户更信结果一致的推理，即使逻辑有误也持续依赖。
- 自信语气会抑制纠错意识，适合关注可信解释设计的研究者。
解释常被视为提升透明度的工具，但也可能助长确认偏见；用户往往在输出看似合理时，默认其推理正确。本文通过系统性扰动推理链并操控表达语气，研究视觉语言模型（VLMs）在多模态道德场景中链式思维（CoT）解释的双重作用。研究发现：（1）用户常将信任与结果一致性挂钩，即使推理存在错误仍持续依赖；（2）自信语气会压制错误识别能力，同时维持信任，表明表达风格可压倒逻辑正确性。结果揭示了CoT解释既能澄清又可能误导的双刃剑特性，强调NLP系统应设计能促进批判性思考而非盲信的解释方式。所有代码将公开发布。
原文摘要 · Abstract (English)
Explanations are often promoted as tools for transparency, but they can also foster confirmation bias; users may assume reasoning is correct whenever outputs appear acceptable. We study this double-edged role of Chain-of-Thought (CoT) explanations in multimodal moral scenarios by systematically perturbing reasoning chains and manipulating delivery tones. Specifically, we analyze reasoning errors in vision language models (VLMs) and how they impact user trust and the ability to detect errors. Our findings reveal two key effects: (1) users often equate trust with outcome agreement, sustaining reliance even when reasoning is flawed, and (2) the confident tone suppresses error detection while maintaining reliance, showing that delivery styles can override correctness. These results highlight how CoT explanations can simultaneously clarify and mislead, underscoring the need for NLP systems to provide explanations that encourage scrutiny and critical thinking rather than blind trust. All code will be released publicly.
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。