Hello! I’m a Senior Staff Research Scientist at Google DeepMind in New York working on core Gemini modeling.
I currently lead an amazing team of six research scientists working on RL scaling techniques, multi-agent RL, and self-improvement. During my tenure, my contributions to Gemini have spanned pretraining data & architectures, domain-specific mid-training, core post-training for Gemini releases, and modeling for Gemini’s IMO 2025 gold medal.
Prior to Gemini, I co-invented generative retrieval, in particular: semantic identifiers, an approach now further refined and widely adopted in generative search and recommendation systems such as Spotify’s. I also pioneered one of the earliest large-scale deployments of efficient character-level transformers at Google, and made core contributions to pretraining objectives research.
Prior to research, I worked on a wide variety of projects across Google and Google Research including planet-scale distributed storage systems, visual analytics, misinformation, fact checking, and news applications. Before that I was a full-stack web developer, and before that I attended Brown University as an undergraduate in 2016. :)
Selected Publications
Advanced version of Gemini with Deep Think officially achieves gold-medal standard at the International Mathematical Olympiad (2025)
Gemini IMO Team (Key Contributor)
DeepMind Blog
Gemini 2.5: Pushing the frontier with advanced reasoning, multimodality, long context, and next generation agentic capabilities (2025)
Gemini Team
arXiv
BIG-Bench Extra Hard (2025)
Mehran Kazemi, Bahare Fatemi, Hritik Bansal, John Palowitch, Chrysovalantis Anastasiou, Sanket Vaibhav Mehta, Lalit K. Jain, Virginia Aglietti, Disha Jindal, Peter Chen, Nishanth Dikkala, Gladys Tyen, Xin Liu, Uri Shalit, Silvia Chiappa, Kate Olszewska, Yi Tay, Vinh Q. Tran, Quoc V. Le, Orhan Firat
ACL 2025
Smaller, Weaker, Yet Better: Training LLM Reasoners via Compute-Optimal Sampling (2024)
Hritik Bansal, Arian Hosseini, Rishabh Agarwal, Vinh Q. Tran, Seyed Mehran Kazemi
ICLR 2025
Recommender Systems with Generative Retrieval (2023)
Shashank Rajput, Nikhil Mehta, Anima Singh, Raghunandan Hulikal Keshavan, Trung Vu, Lukasz Heldt, Lichan Hong, Yi Tay, Vinh Q. Tran, Jonah Samost, Maciej Kula, Ed Chi, Maheswaran Sathiamoorthy
NeurIPS 2023
Transcending scaling laws with 0.1% extra compute (2022)
Yi Tay, Jason Wei, Hyung Won Chung, Vinh Q. Tran, David R. So, Siamak Shakeri, Xavier Garcia, Huaixiu Steven Zheng, Jinfeng Rao, Aakanksha Chowdhery, Denny Zhou, Donald Metzler, Slav Petrov, Neil Houlsby, Quoc V. Le, Mostafa Dehghani
EMNLP 2023
Transformer Memory as a Differentiable Search Index (2022)
Yi Tay*, Vinh Q. Tran*, Mostafa Dehghani, Jianmo Ni, Dara Bahri, Harsh Mehta, Zhen Qin, Kai Hui, Zhe Zhao, Jai Gupta, Tal Schuster, William W. Cohen, Donald Metzler
NeurIPS 2022
UL2: Unifying Language Learning Paradigms (2022)
Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Xavier Garcia, Dara Bahri, Tal Schuster, Huaixiu Steven Zheng, Neil Houlsby, Donald Metzler
ICLR 2023
A New Generation of Perspective API: Efficient Multilingual Character-level Transformers (2022)
Alyssa Lees*, Vinh Q. Tran*, Yi Tay*, Jeffrey Sorensen, Jai Gupta, Donald Metzler, Lucy Vasserman
KDD 2022 ADS
Charformer: Fast Character Transformers via Gradient-based Subword Tokenization (2021)
Yi Tay*, Vinh Q. Tran*, Sebastian Ruder, Jai Gupta, Hyung Won Chung, Dara Bahri, Zhen Qin, Simon Baumgartner, Cong Yu, and Donald Metzler.
ICLR 2022
Please see my full list and recent publications on my Scholar.
Mentoring & Service
Mentored Student Researchers & Interns @ Google
- Hritik Bansal (2024 Student Researcher, PhD @ UCLA)
- Ronak Pradeep (2023 Student Researcher, PhD @ Waterloo)
- Sanket Vaibhav Mehta (2022 Research Intern, now RS on my team!)
- Yuanzhe (Richard) Pang (2020 Research Intern, now RS @ Meta)
- Kate Lin (2019 STEP Intern, now SWE @ Google Research)
- Amy Pu (2019 STEP Intern, now SWE @ YouTube Music Recommendations)
- Daniil Dmitriev (2017 SWE Intern, now PhD @ ETH Zurich)
Reviewer for NeurIPS, ICLR, SIGIR, etc.