# LoGU: Long-form Generation with Uncertainty Expressions Source: [LoGU: Long-form Generation with Uncertainty Expressions]() Collection index: [llms.txt]() Data updated: 2026-01-04T04:57:32.958Z Authors: Ruihan Yang, Caiqi Zhang, Zhisong Zhang, Xinting Huang, Sen Yang, Nigel Collier, Dong Yu, Deqing Yang Published: 2024-10-18T09:15:35.000Z ## Abstract While Large Language Models (LLMs) demonstrate impressive capabilities, they still struggle with generating factually incorrect content (i.e., hallucinations). A promising approach to mitigate this issue is enabling models to express uncertainty when unsure. Previous research on uncertainty modeling has primarily focused on short-form QA, but realworld applications often require much longer responses. In this work, we introduce the task of Long-form Generation with Uncertainty(LoGU). We identify two key challenges: Uncertainty Suppression, where models hesitate to express uncertainty, and Uncertainty Misalignment, where models convey uncertainty inaccurately. To tackle these challenges, we propose a refinement-based data collection framework and a two-stage training pipeline. Our framework adopts a divide-and-conquer strategy, refining uncertainty based on atomic claims. The collected data are then used in training through supervised fine-tuning (SFT) and direct preference optimization (DPO) to enhance uncertainty expression. Extensive experiments on three long-form instruction following datasets show that our method significantly improves accuracy, reduces hallucinations, and maintains the comprehensiveness of responses. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Factuality and Uncertainty Estimation on Bios (test)](): Long-form generation factuality and uncertainty estimation; Bios (test); FA: 71.4. - [Long-form generation factuality and uncertainty estimation on LongFact (test)](): Long-form generation factuality and uncertainty estimation; LongFact (test); Factuality Score: 91.5. - [Long-form generation factuality and uncertainty estimation on WildHallu (test)](): Long-form generation factuality and uncertainty estimation; WildHallu (test); Factuality Score: 0.86. ## Original sources and implementation [Original paper]() [PDF]() [Implementation]()