Frontier Data Summit 2026 lineup reveals a dozen new AI benchmarks and top researchers

dlwh · x · 2026-09-22

Snorkel AI shared details of the invite-only Frontier Data Summit (Oct 8, 2026, San Francisco) focused on AI benchmarks and evaluation research. Announced benchmarks include STELLA-Bench, MedPAIR, FutureSim, CollusionBench, HalluWorld, JudgmentBench, ORCA-Bench, and Terminal Bench 2.1, with topics spanning agent evaluation in realistic environments, autonomy horizons, multi-artifact output scoring, and train-to-test scaling laws. Speakers include François Chollet (ARC Prize), Dawn Song (UC Berkeley), Sanmi Koyejo (Stanford), Alex Ratner (Snorkel AI), and Alex Shaw (Terminal Bench), plus a fireside on how benchmarks are actually used in model development with Laude Institute's Braden Hancock.

Original post →

More from Companies & People

Companies & People channel →