ICML 2026 — live archive
Running archive of ICML 2026 — notes captured live and expanded afterward. Standout papers get promoted to full reading-archive entries; this page is the day-by-day index.
Accuracy rule (per the blog’s workflow): real titles, authors, and numbers only. Anything not yet verified is marked [unverified] until checked against the source.
The Forty-Third International Conference on Machine Learning — Seoul, South Korea, COEX Convention & Exhibition Center, July 6–11, 2026. Structure: July 6 = Expo & Tutorials · July 7–9 = Main Conference · July 10–11 = Workshops.
Program
July 6 — Expo & Tutorials
July 7 — Main Conference (Day 1)
Invited Talk — Causal Inference with Transformer Models Susan Athey (Stanford GSB) · Hall C · 4:00–5:00 PM KST
Answering causal questions about sequence data — text, career/job sequences, customer journeys. Methods for average treatment effects, conditional ATEs, and decompositions of across-group differences in average outcomes, under both experimental and observational data.
Speaker: Susan Athey — The Economics of Technology Professor, Stanford GSB; 2023 President of the American Economic Association; Chief Economist, U.S. DOJ Antitrust Division (2022–2024); founding director, Golub Capital Social Impact Lab.
Source: ICML 2026 virtual schedule. Talk content as listed; my notes to be added.