# SpanUQ: Span-Level Uncertainty Quantification for Large Language Model Generation Source: [SpanUQ: Span-Level Uncertainty Quantification for Large Language Model Generation]() Collection index: [llms.txt]() Data updated: 2026-07-09T05:00:01.225Z Authors: Yimeng Zhang, Yingying Zhuang, Ziyi Wang, Yuxuan Lu, Pei Chen, Aman Gupta, Zhe Su, Ming Tan, Zhilin Zhang, Qun Liu, Manikandarajan Ramanathan, Rajashekar Maragoud, Edward Vul, Jing Huang, Dakuo Wang Published: 2026-07-07T01:09:46.000Z ## Abstract Uncertainty estimation is essential not only for the trustworthy deployment of large language models (LLMs) but also as a foundation for self-refinement in LLM generation. However, existing approaches operate at suboptimal granularities: token-level scores lack semantic coherence, while sequence-level scores fail to localize errors. We formalize Span-Level Uncertainty Estimation (SLUE), a new task that targets the natural granularity for uncertainty: semantically coherent text spans, each conveying a single assessable unit of meaning. To address this task, we introduce SPANUQ, a lightweight probe that distills the uncertainty knowledge from expensive multi-sample inference into a single forward pass over LLM hidden states. SPANUQ employs a DETR-style span decoder to simultaneously detect spans and estimate their uncertainty via a Mixture of Beta distribution, trained with a principled combination of Beta NLL regression and contrastive ranking objectives. We construct SPANUQ-BENCH, the first span-level uncertainty benchmark comprising 20K prompts, 293K annotated spans, and continuous soft labels derived from multi-sample claim verification. Experiments on five LLM backbones show that SPANUQ consistently achieves the best span-level uncertainty quality, outperforming the strongest probe baseline and all sampling-based methods while being 10-20x faster. Its DETR-based span detector attains 0.910 F1, surpassing the best heuristic by 39.4%, enabling precise error localization that sequence-level methods cannot provide. The framework generalizes across five LLMs spanning two model families. ## 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. - [Span-level uncertainty estimation on SPANUQ-BENCH v1.0 (test)](): Span-level uncertainty estimation; SPANUQ-BENCH v1.0 (test); AUROC: 0.944. - [Sequence-level uncertainty estimation on SPANUQ-BENCH 2,000 sequences (test)](): Sequence-level uncertainty estimation; SPANUQ-BENCH 2,000 sequences (test); Spearman Correlation: 0.851. ## Original sources and implementation [Original paper]() [PDF]()