Scholar
Abhinav Gupta
Google Scholar ID: jAaCd7YAAAAJ
Research Scientist, Google DeepMind
Citations & Impact
Citations
870
H-index
10
i10-index
10
Publications
20
Co-authors
0
Publications
Resume (English only)
Academic Achievements
- - Thesis: 'Self-play for human-agent communication'
- - Paper: 'Dynamic population-based meta-learning for multi-agent communication with natural language', NeurIPS 2021
- - Paper: 'On the interaction between supervision and self-play in emergent communication', ICLR 2020
- - Paper: 'Capacity, Bandwidth, and Compositionality in Emergent Language Learning', AAMAS 202
Research Experience
- - Google DeepMind, London, advised by Grzegorz Swirszcz, Wojciech Czarnecki, and Oriol Vinyals, working on opponent behavior modeling
- - Meta AI, Seattle, advised by Madian Khabsa and Roberta Raileanu, improving generalization in RL using uncertainty estimates
- - Microsoft Research NYC, advised by Jordan Ash, replacing PPO with weighted SFT in RLHF
- - Google Research Labs team, Mountain View, advised by Navneet Potti, enhancing code correctness in LLMs through RL with execution feedback
- - Microsoft Research Cambridge, working with Raluca Georgescu, Sam Devlin, and Katja Hofmann, learning human-like behavior using offline RL and imitation learning
Education
- Ph.D. from the Department of Computer Science and Operations Research at the University of Montreal, advised by Christopher Pal; Master's degree from the Courant Institute of Mathematical Sciences at New York University.
Background
- Research interests include the effects of self-play in emergent languages arising in multi-agent communication systems, with a focus on computer science and operations research.
Miscellany
- Personal interests not mentioned
Co-authors