Four universities share the lead with two major AI paper awards each
Across a defined set of 11 flagship AI conferences from 2024 through 20 August 2026, ETH Zurich, Oxford, Toronto and Washington each appear on two verified Best Paper or Outstanding Paper winners. Counts credit every university listed as a primary author affiliation and exclude honorable mentions, runners-up and workshop awards.
2
winning papers per leader
4
universities tied for first
8
leader-linked papers
11
flagship venues surveyed
Verified award-winning papers
Blue = papers carrying the university affiliation
What separates the leaders
01
Oxford repeats in computer vision. Its two wins are consecutive CVPR Best Papers, VGGT in 2025 and D4RT in 2026.
02
Washington spans language and general AI. It appears on EMNLP 2025’s Best Paper and a NeurIPS 2025 Best Paper.
03
ETH combines theory across NLP and learning. Its authors appear on ACL 2026’s Best Paper and an ICLR 2026 Outstanding Paper.
04
Toronto’s wins cross AAAI and ACL. One is an AAAI 2025 Outstanding Paper and the other an ACL 2026 Outstanding Paper.
The eight winning papers
| University | Paper | Venue | Award | Primary author affiliations reported |
|---|---|---|---|---|
| ETH Zurich | Characterizing the Expressivity of Local Attention in Transformers | ACL 2026 | Best Paper | ETH Zurich |
| ETH Zurich | Transformers Are Inherently Succinct | ICLR 2026 | Outstanding Paper | RPTU Kaiserslautern-Landau; ETH Zurich; Max Planck Institute for Software Systems |
| University of Oxford | VGGT: Visual Geometry Grounded Transformer | CVPR 2025 | Best Paper | University of Oxford; Meta AI |
| University of Oxford | Efficiently Reconstructing Dynamic Scenes One D4RT at a Time | CVPR 2026 | Best Paper | Google DeepMind; University College London; University of Oxford |
| University of Toronto | Every Bit Helps: Achieving the Optimal Distortion with a Few Queries | AAAI 2025 | Outstanding Paper | University of Toronto |
| University of Toronto | Systematicity between Forms and Meanings across Languages Supports Efficient Communication | ACL 2026 | Outstanding Paper | Saarland University; University of Toronto |
| University of Washington | Infini-gram Mini: Exact n-gram Search at the Internet Scale with FM-Index | EMNLP 2025 | Best Paper | University of Washington; Allen Institute for AI; Stanford University |
| University of Washington | Artificial Hivemind: The Open-Ended Homogeneity of Language Models (and Beyond) | NeurIPS 2025 | Best Paper | University of Washington; Allen Institute for AI; Carnegie Mellon University; Stanford University; Lila Sciences |
Method: live-web scan of AAAI, IJCAI, ICLR, ICML, NeurIPS, ACL, EMNLP, NAACL, CVPR, ICCV and ECCV award and paper pages, covering 2024 through 20 August 2026. A paper counts once for every university named in its authors’ primary affiliations; corporate and independent institutes are shown but not ranked. Main-conference Best/Outstanding awards only.