Talk: Toward Reliable Deployment of LLM Agents via Uncertainty Quantification

SharonYixuanLi · x · 2026-09-01

The AI Agent Frontier Seminar will feature Prof. Sharon Li (UW-Madison, ICML 2026 Chair). Her talk, "Toward Reliable Deployment of LLM Agents," will explore how to detect silent failures in long-horizon tasks using Uncertainty Quantification (UQ) and the "progress advantage" signal derived from RL post-training.

Original post →

More from Companies & People

Companies & People channel →