用自述理由建坐标系，分析大模型决策是否靠谱。
LAMP: Extracting Local Decision Surfaces From Large Language Models
- 以模型自述解释为坐标，拟合局部线性代理模型
- 在多个任务中发现模型决策与人类判断高度一致
- 无需梯度或内部数据，适合审计闭源模型
我们提出 LAMP（Local Attribution Mapping Probe），一种揭示黑箱语言模型决策面的方法，通过近似决策表面来研究模型将其陈述原因映射到预测结果的可靠性。LAMP 将模型自身的自述解释作为坐标系，拟合一个局部线性代理模型，将这些解释权重与模型输出关联。该方法揭示了陈述因素对模型决策的影响程度。我们在情感分析、争议话题检测和安全提示审计三个任务上应用 LAMP。结果显示，大多数情况下，模型的局部线性决策景观与人类对解释质量的判断一致，并在临床病历数据集上与专家评估相符。由于 LAMP 不需要访问模型梯度、logits 或内部激活，因此可作为实用且轻量的框架，用于审计专有语言模型，评估其行为是否与其提供的解释保持一致。
原文摘要 · Abstract (English)
We introduce LAMP (Local Attribution Mapping Probe), a method that shines light onto a black-box language model's decision surface and studies how reliably a model maps its stated reasons to its reported predictions by approximating a decision surface. LAMP treats the model's own self-reported explanations as a coordinate system and fits a locally linear surrogate that links those weights to the model's output. By doing so, it reveals how much the stated factors steer the model's decisions. We apply LAMP to three tasks: sentiment analysis, controversial-topic detection, and safety-prompt auditing. Across these tasks, LAMP reveals that many language models' locally approximated linear decision landscapes overall agree with human judgments on explanation quality and, on a clinical case-file data set, align with expert assessments. Since LAMP operates without requiring access to model gradients, logits, or internal activations, it serves as a practical and lightweight framework for auditing proprietary language models, and enabling assessment of whether a model appears to behave consistently with the explanations it provides.
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。