通过检测模型内部表示对扰动的稳定性，提升大模型预测可信度。
Calibrating LLM Confidence by Probing Perturbed Representation Stability
- 用对抗性扰动测试模型隐藏状态的响应稳定性，提取鲁棒特征。
- 在多个大模型上使校准误差降低55%，准确率提升5个百分点。
- 无需训练、适配性强，适合需高可信度的应用场景。
大语言模型（LLMs）的置信度失准影响其可靠性，亟需精准的置信度估计。本文提出CCPS（Calibrating LLM Confidence by Probing Perturbed Representation Stability），通过在最终隐藏状态施加定向对抗扰动，提取模型对扰动的响应特征，并用轻量分类器预测答案正确性。在8B至32B参数的Llama、Qwen、Mistral等模型上，使用MMLU和MMLU-Pro基准进行多选与开放题测试。结果表明，相较于最强现有方法，CCPS在四个LLM和三个MMLU变体上，将期望校准误差降低约55%，布里尔得分减少21%，准确率提升5个百分点，精确率-召回率曲线下面积增加4个百分点，受试者工作特征曲线下面积提高6个百分点。该方法高效、普适且更精准，显著增强模型可信度。
原文摘要 · Abstract (English)
Miscalibration in Large Language Models (LLMs) undermines their reliability, highlighting the need for accurate confidence estimation. We introduce CCPS (Calibrating LLM Confidence by Probing Perturbed Representation Stability), a novel method analyzing internal representational stability in LLMs. CCPS applies targeted adversarial perturbations to final hidden states, extracts features reflecting the model's response to these perturbations, and uses a lightweight classifier to predict answer correctness. CCPS was evaluated on LLMs from 8B to 32B parameters (covering Llama, Qwen, and Mistral architectures) using MMLU and MMLU-Pro benchmarks in both multiple-choice and open-ended formats. Our results show that CCPS significantly outperforms current approaches. Across four LLMs and three MMLU variants, CCPS reduces Expected Calibration Error by approximately 55% and Brier score by 21%, while increasing accuracy by 5 percentage points, Area Under the Precision-Recall Curve by 4 percentage points, and Area Under the Receiver Operating Characteristic Curve by 6 percentage points, all relative to the strongest prior method. CCPS delivers an efficient, broadly applicable, and more accurate solution for estimating LLM confidence, thereby improving their trustworthiness.
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。