让大模型学会逐条判断自己回答中每个事实的可信度并自我修正。
Fact-Level Confidence Calibration and Self-Correction
- 按事实粒度校准信心，结合问题相关性加权评估正确性。
- 在四个数据集上显著减少幻觉，提升低置信度事实准确性。
- 无需外部知识源，适合需要高可信输出的场景如医疗、金融。
大模型的置信度校准旨在使其自评信心与实际回答准确率对齐，从而实现对输出正确性的自我评估。然而，现有校准方法通常仅估算两个标量来代表整体响应的信心与正确性，这在长文本生成中不足——因回应包含多个独立事实，可能部分正确、部分存疑。同时，这些方法忽略了每个事实与问题的相关性。为此，我们提出事实级校准框架（Fact-Level Calibration），在更细粒度上实现基于相关性加权的正确性校准。该框架的深入分析进一步催生了置信度引导的事实级自我修正方法（ConFix），利用高置信度事实作为额外知识来改进低置信度事实。在四个数据集和六种模型上的大量实验表明，ConFix能有效缓解幻觉现象，且无需依赖检索系统等外部知识源。
原文摘要 · Abstract (English)
Confidence calibration in LLMs, i.e., aligning their self-assessed confidence with the actual accuracy of their responses, enabling them to self-evaluate the correctness of their outputs. However, current calibration methods for LLMs typically estimate two scalars to represent overall response confidence and correctness, which is inadequate for long-form generation where the response includes multiple atomic facts and may be partially confident and correct. These methods also overlook the relevance of each fact to the query. To address these challenges, we propose a Fact-Level Calibration framework that operates at a finer granularity, calibrating confidence to relevance-weighted correctness at the fact level. Furthermore, comprehensive analysis under the framework inspired the development of Confidence-Guided Fact-level Self-Correction ($\textbf{ConFix}$), which uses high-confidence facts within a response as additional knowledge to improve low-confidence ones. Extensive experiments across four datasets and six models demonstrate that ConFix effectively mitigates hallucinations without requiring external knowledge sources such as retrieval systems.
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。