让大模型单次推理生成多种思路，快速估算回答可信度。
Dist2ill: Distributional Distillation for One-Pass Uncertainty Estimation in Large Language Models
- 用分布蒸馏训练模型一次推理生成多种解题路径
- 在保持思路多样性的同时，将校准误差降低37.6%
- 适合需要快速可靠置信度评估的部署场景
大型语言模型常出现生成质量与自身置信度不匹配的问题。贝叶斯方法虽能有效解决，但需多次采样，计算开销大。为实现单次前向传播下的高精度不确定性估计，我们提出分布蒸馏框架Dist2ill：训练模型在一次推理中生成多个多样化的推理路径，并通过轻量级参数模块逼近从采样分布中得出的经验置信度。大量实验表明，Dist2ill在保持推理多样性的同时，显著提升校准性能，预期校准误差（ECE）和负对数似然（NLL）均达当前最优水平，且计算高效。
原文摘要 · Abstract (English)
Large Language Models (LLMs) often exhibit misalignment between the quality of their generated responses and the confidence estimates they assign to them. Bayesian treatments, such as marginalizing over a reliable weight posterior or over the space of reasoning traces, provide an effective remedy, but incur substantial computational overhead due to repeated sampling at test time. To enable accurate uncertainty estimation in a single forward pass, we propose a novel distributional distillation framework (Dist2ill) that trains an LLM to produce multiple diverse reasoning paths within one inference pass, while using a lightweight parametric module to approximate empirical confidence scores derived from the sampling distribution. Extensive experiments demonstrate that Dist2ill preserves reasoning diversity and achieves state-of-the-art uncertainty estimation, substantially improving Expected Calibration Error (ECE) and Negative Log-Likelihood (NLL), while remaining computationally efficient.
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。