Detecting Uncertainty to Enhance AI Reliability @SimonsInstitute
Detecting Uncertainty to Enhance AI Reliability  @SimonsInstitute
Uploaded October 2026 | Updated November 2026, 2 days ago
Litian Liu (Qualcomm AI research)
https://simons.berkeley.edu/talks/litian-liu-qualcomm-ai-research-2026-10-08
Trustworthy AI: From Hallucinations to Reliable Autonomy

Detecting hallucinations in large language models is a critical open problem with significant implications for safety and reliability. While existing hallucination detection methods achieve strong performance in question-answering tasks, they remain less effective on tasks requiring reasoning. In this work, we revisit hallucination detection through the lens of out-of-distribution (OOD) detection, a well-studied problem in areas like computer vision. Treating next-token prediction in language models as a classification task allows us to apply OOD techniques, provided appropriate modifications are made to account for the structural differences in large language models. We show that OOD-based approaches yield training-free, single-sample-based detectors, achieving strong accuracy in hallucination detection for reasoning tasks. Overall, our work suggests that reframing hallucination detection as OOD detection provides a promising and scalable pathway toward language model safety.
Detecting Uncertainty to Enhance AI ReliabilityQuantum ergodicity on graphs (Part 1)Extractors for Samplable Distributions from the Two-Source Extractor RecipeStrong convergence of right-angled Artin groups and spectral gaps of hyperbolic 3-manifoldsWeak Poincaré Inequalities and Mixing from Non-Worst-Case InitializationsDo we need diffusion in robotics?On Spectral Independence and Localization Schemes (Part 2)Learning the structure of open quantum systemsNormalizing Flow MapsStrong Convergence (Part 3)High-Dimensional Expansion (HDX) (Part 2)The Pseudorandomness Toolkit (Part 3)
Simons Institute for the Theory of Computing |

Detecting Uncertainty to Enhance AI Reliability

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER