AI System Achieves Breakthrough in Mathematical Olympiad, Securing Official Recognition
An artificial intelligence system has successfully submitted complete and correct solutions to teh International Mathematical Olympiad (IMO) 2025, marking a significant advancement in AI’s capability in complex problem-solving. While the IMO has confirmed the accuracy of the submitted answers, their review did not extend to the system’s underlying technology or processes.
The advancement of the IMO 2025 system involved a collaborative effort across multiple teams. Dawsen Hwang and vincent Cohen-Addad co-led an enhanced inference approach, while Dawsen Hwang also led the Deep Think inference component. Key contributions to modeling came from Theophane Weber and Ankesh Anand. The inference process benefited from the work of Vinay Ramasesh,Andreas Kirsch,Jieming Mao,Zicheng Xu,Wilfried Bounsi,and Vahab Mirrokni. Training data was curated by Hoang Nguyen, Fred Zhang, Mahan Malihi, and Yangsibo Huang.
Significant contributions also came from related teams. The AlphaGeometry team, led by Yuri Chervonyi, included Trieu Trinh, Hoang Nguyen, Junsu Kim, Mirek Olšák, Marcelo Menegali, and Xiaomeng Yang. Formal mathematics expertise was provided by Miklós Z. Horváth, Aja Huang, and Goran Žužić. Support and collaboration were extended by fabian Pedregosa, Richard Song, Alex Zhai, Sara Javanmardi, YaGuang Li, Filipe Miguel de Almeida, Silvio Lattanzi, Ashkan Norouzi Fard, Tal Schuster, Honglu Fan, Xuezhi Wang, Aditi Mavalankar, Tom Schaul, and Rosemary Ke.
Core members of the Deep Think team,including Archit Sharma,Tong He,and Shubha Raghvendra,were instrumental. The post-training phase saw contributions from Tianhe Kevin Yu, siamak Shakeri, Hanzhao Lin, Cosmo Du, and Sid Lall. The research built upon foundational work in the Thinking Area.
The project was guided by advisors Quoc Le and Pushmeet Kohli, with program support from Kristen Chiafullo and Alex Goldin.
A dedicated group of experts provided crucial data and evaluations. Insuk Seo led this effort, which included Jiwon Kang, Donghyun kim, Junsu Kim, Jimin Kim, Seongbin Jeon, Yoonho Na, Seunghwan Lee, Jihoo Lee, Younghun Jo, Yongsuk Hur, Seongjae Park, Kyuhyeon choi, Minkyu Choi, Su-Hyeok Moon, Seojin Kim, Yueun Lee, Taehun Kim, jeeho Ryu, Seungwoo Lee, Dain Kim, Sanha Lee, Hyunwoo Choi, Aiden Jung, Youngbeom Jin, Jeonghyun Ahn, Junhwi Bae, Gyumin Kim, Nam Dung tran, Cheng-Chiang Tsai, kari Ragnarsson, Kiat Chuan Tan, Yahya Tabesh, Hamed Mahdavi, Azin Nazari, Xiangzhuo ding, Chu-Lan Kao, Steven Creech, Tony Feng, and Ciprian Manolescu.
Further support was provided by Jessica Lo and Sajjad zafar for compute resources, and Jane Labanowski, Andy Forbes, and Sean Nakamoto for legal and logistics. Advice and support were also received from Omer Levy, Timothy Lillicrap, jack Rae, Yifeng Lu, Heng-tze Cheng, Ed Chi, Vahab Mirrokni, Tulsee Doshi, Madhavi Sewak, Melvin Johnson, Koray Kavukcuoglu, Oriol Vinyals, Jeff Dean, Demis hassabis, and Sergey Brin.
Professor Gregor Dolinar from the IMO Board offered support and endorsement for the project.
The IMO’s confirmation of the submitted answers’ completeness and correctness signifies a milestone in AI’s application to advanced mathematical challenges. The IMO’s official statement notes that their review process does not encompass the validation of the AI system, its methodologies, or its underlying model.
