EUVIMEDEUVIMED entdecken
Crossref

Sampling-Free Uncertainty Quantification via Hidden State Dynamics in Language Models

Yixin Bu, Guanyun Zou, Renzhi Wang, Runze Xia, Cunjun Wang, Hongliang Dai, Xiaoqing Ma, Piji Li

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (36) · 30104-30111

DOI: 10.1609/aaai.v40i36.40259 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Yixin Bu und 7 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Large language models (LLMs) demonstrate remarkable capabilities in various complex language tasks, yet they face significant reliability challenges, including factual inaccuracies and generated biases. Uncertainty quantification (UQ) plays a pivotal role in assessing model trustworthiness, particularly for high-stakes applications. However, current UQ methods for LLMs encounter computational efficiency bottlenecks due to their reliance on extensive sampling or external model invocations. In this work, we introduce a novel, sampling-free uncertainty quantification framework centered on hidden layer representation analysis. Our method facilitates real-time uncertainty quantification by modeling hierarchical internal semantic dynamics during the generation process. Through comprehensive experiments on multiple QA datasets and diverse model scales, we show that our approach consistently outperforms existing uncertainty quantification techniques in distinguishing correct from incorrect generations. Our results reveal that analyzing the dynamic evolution of hidden states provides a potent and computationally efficient signal for uncertainty quantification, directly from the model's internal workings, surpassing methods that depend solely on output probabilities or approximations via multiple samples.

Zitieren

APA

Bu, Y., Zou, G., Wang, R., Xia, R., Wang, C., Dai, H., Ma, X., & Li, P. (2026). Sampling-Free Uncertainty Quantification via Hidden State Dynamics in Language Models. Proceedings of the AAAI Conference on Artificial Intelligence, 40(36), 30104-30111. https://doi.org/10.1609/aaai.v40i36.40259

Vancouver

Bu Y, Zou G, Wang R, Xia R, Wang C, Dai H, et al. Sampling-Free Uncertainty Quantification via Hidden State Dynamics in Language Models. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(36):30104-30111. doi:10.1609/aaai.v40i36.40259

BibTeX
@article{Bu2026,
  author = {Bu, Yixin and Zou, Guanyun and Wang, Renzhi and Xia, Runze and Wang, Cunjun and Dai, Hongliang and Ma, Xiaoqing and Li, Piji},
  title = {Sampling-Free Uncertainty Quantification via Hidden State Dynamics in Language Models},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {36},
  pages = {30104--30111},
  doi = {10.1609/aaai.v40i36.40259}
}
Kostenlos starten →