用结构化反馈修正大模型长文本事实错误，无需重训练
FactCorrector: A Graph-Inspired Approach to Long-Form Factuality Correction of Large Language Models
- 基于图结构设计反馈机制，动态修正事实性错误
- 在VELI5等数据集上事实准确率显著提升，相关性保持不变
- 适合需要高可信度生成的场景，如医疗、金融问答
大语言模型在知识密集型应用中广泛使用，但常产生事实错误。本文提出FactCorrector，一种无需重训练即可跨领域适配的后处理修正方法，利用原始回答的事实性结构化反馈生成修正结果。为支持严谨评估，我们构建了VELI5基准数据集，包含系统注入的事实错误及真实修正答案。在VELI5及多个主流长文本事实性数据集上的实验表明，该方法显著提升事实精确率，同时保持内容相关性，优于多个强基线模型。代码已公开于https://ibm.biz/factcorrector。
原文摘要 · Abstract (English)
Large language models (LLMs) are widely used in knowledge-intensive applications but often generate factually incorrect responses. A promising approach to rectify these flaws is correcting LLMs using feedback. Therefore, in this paper, we introduce FactCorrector, a new post-hoc correction method that adapts across domains without retraining and leverages structured feedback about the factuality of the original response to generate a correction. To support rigorous evaluations of factuality correction methods, we also develop the VELI5 benchmark, a novel dataset containing systematically injected factual errors and ground-truth corrections. Experiments on VELI5 and several popular long-form factuality datasets show that the FactCorrector approach significantly improves factual precision while preserving relevance, outperforming strong baselines. We release our code at https://ibm.biz/factcorrector.
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。