The ICML 2026 awards
Both of the conference's outstanding papers are about diffusion models; the Test of Time award went to a 2016 paper on asynchronous reinforcement learning.
Why it matters
At the largest machine learning conference both top papers of the year concern diffusion models, and one of them questions the assumed benefit of arbitrary generation order.
The announcement is dated 5 July 2026. The committee recognised two Outstanding Papers and five Honorable Mentions. The first is The Flexibility Trap: Rethinking the Value of Arbitrary Order in Diffusion Language Models by Zanlin Ni, Shenzhi Wang, Yang Yue, Tianyu Yu, Weilin Zhao, Yeguo Hua, Tianyi Chen, Jun Song, Cheng Yu, Bo Zheng and Gao Huang. The second is High-Accuracy Sampling for Diffusion Models and Log-Concave Distributions by Fan Chen, Sinho Chewi, Constantinos Daskalakis and Alexander Rakhlin. The Test of Time Award went to the ICML 2016 paper Asynchronous Methods for Deep Reinforcement Learning by Volodymyr Mnih, Adria Puigdomenech Badia, Mehdi Mirza, Alex Graves, Timothy P. Lillicrap, Tim Harley, David Silver and Koray Kavukcuoglu.