Acknowledgements
We thank the Worldwide Mathematical Olympiad group for his or her help.
Thang Luong led the general technical course of the superior Gemini mannequin with Deep Suppose for IMO and co-led with Edward Lockhart on the general coordination of the IMO 2025 effort.
The IMO 2025 system wouldn’t have been attainable with out the next technical leads. Dawsen Hwang, Junehyuk Jung co-led coaching information and skilled analysis. Jonathan Lee, Nate Kushman, Pol Moreno, Yi Tay co-led the coaching of the superior Gemini Deep Suppose mannequin whereas Lei Yu led mannequin analysis. Golnaz Ghiazi, Garrett Bingham, Lalit Jain co-led Deep Suppose inference whereas Dawsen Hwang, Vincent Cohen-Addad co-led an enhanced inference strategy.
The IMO 2025 system was additionally developed with key contributions from Theophane Weber, Ankesh Anand for modeling; Vinay Ramasesh, Andreas Kirsch, Jieming Mao, Zicheng Xu, Wilfried Bounsi, Vahab Mirrokni for inference; Hoang Nguyen, Fred Zhang, Mahan Malihi, Yangsibo Huang for coaching information.
We thank contributions from associated groups and efforts. AlphaGeometry crew with Yuri Chervonyi (lead), Trieu Trinh, Hoang Nguyen, Junsu Kim, Mirek Olšák, Marcelo Menegali, Xiaomeng Yang. Miklós Z. Horváth, Aja Huang, Goran Žužić for formal arithmetic. We thank Fabian Pedregosa, Richard Music, 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, Rosemary Ke for help and collaboration.
We particularly thank different core members of the Deep Suppose crew (Archit Sharma, Tong He, Shubha Raghvendra), the post-training effort (Tianhe Kevin Yu, Siamak Shakeri, Hanzhao Lin, Cosmo Du, Sid Lall), and Pondering Space analysis that the IMO 2025 system have been constructed on.
This effort was suggested by Quoc Le and Pushmeet Kohli, with program help from Kristen Chiafullo and Alex Goldin.
We’d additionally prefer to thank our consultants for offering information and evaluations: Insuk Search engine optimization (lead), 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, Ciprian Manolescu.
Additional because of Jessica Lo and Sajjad Zafar for his or her help for compute provision and administration; Jane Labanowski, Andy Forbes, Sean Nakamoto for authorized and logistics; and 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 for his or her help and recommendation.
Lastly, we thank Prof Gregor Dolinar from the IMO Board for the help and endorsement.
The IMO have confirmed that our submitted solutions are full and proper options. It is very important notice that their evaluate doesn’t prolong to validating our system, processes, or underlying mannequin (see extra).
Acknowledgements
We thank the Worldwide Mathematical Olympiad group for his or her help.
Thang Luong led the general technical course of the superior Gemini mannequin with Deep Suppose for IMO and co-led with Edward Lockhart on the general coordination of the IMO 2025 effort.
The IMO 2025 system wouldn’t have been attainable with out the next technical leads. Dawsen Hwang, Junehyuk Jung co-led coaching information and skilled analysis. Jonathan Lee, Nate Kushman, Pol Moreno, Yi Tay co-led the coaching of the superior Gemini Deep Suppose mannequin whereas Lei Yu led mannequin analysis. Golnaz Ghiazi, Garrett Bingham, Lalit Jain co-led Deep Suppose inference whereas Dawsen Hwang, Vincent Cohen-Addad co-led an enhanced inference strategy.
The IMO 2025 system was additionally developed with key contributions from Theophane Weber, Ankesh Anand for modeling; Vinay Ramasesh, Andreas Kirsch, Jieming Mao, Zicheng Xu, Wilfried Bounsi, Vahab Mirrokni for inference; Hoang Nguyen, Fred Zhang, Mahan Malihi, Yangsibo Huang for coaching information.
We thank contributions from associated groups and efforts. AlphaGeometry crew with Yuri Chervonyi (lead), Trieu Trinh, Hoang Nguyen, Junsu Kim, Mirek Olšák, Marcelo Menegali, Xiaomeng Yang. Miklós Z. Horváth, Aja Huang, Goran Žužić for formal arithmetic. We thank Fabian Pedregosa, Richard Music, 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, Rosemary Ke for help and collaboration.
We particularly thank different core members of the Deep Suppose crew (Archit Sharma, Tong He, Shubha Raghvendra), the post-training effort (Tianhe Kevin Yu, Siamak Shakeri, Hanzhao Lin, Cosmo Du, Sid Lall), and Pondering Space analysis that the IMO 2025 system have been constructed on.
This effort was suggested by Quoc Le and Pushmeet Kohli, with program help from Kristen Chiafullo and Alex Goldin.
We’d additionally prefer to thank our consultants for offering information and evaluations: Insuk Search engine optimization (lead), 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, Ciprian Manolescu.
Additional because of Jessica Lo and Sajjad Zafar for his or her help for compute provision and administration; Jane Labanowski, Andy Forbes, Sean Nakamoto for authorized and logistics; and 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 for his or her help and recommendation.
Lastly, we thank Prof Gregor Dolinar from the IMO Board for the help and endorsement.
The IMO have confirmed that our submitted solutions are full and proper options. It is very important notice that their evaluate doesn’t prolong to validating our system, processes, or underlying mannequin (see extra).
