Trapit Bansal
Trapit Bansal is an artificial intelligence researcher whose work runs from meta-learning to reasoning models. Among the papers he co-wrote is the ICLR 2018 work Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments. From 2022 he worked at OpenAI, where he helped create the o-series models and worked on reinforcement learning for chain-of-thought reasoning. In June 2025 Meta hired him, and he now works on AI reasoning models at Meta Superintelligence Labs.
- Works
- 31
- Citations
- 1,223
- h-index
- 14
Publication numbers from OpenAlex, as of 15 October 2026.
Timeline
-
2018
Co-author of an ICLR 2018 paper
He co-authors Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments, published at ICLR 2018.
-
2020
Self-supervised meta-learning
He co-writes a self-supervised meta-learning paper for few-shot natural language classification tasks, which is set to appear at EMNLP 2020.
-
2022–2025
Reasoning work at OpenAI
From 2022 he works at OpenAI, and he is a cocreator of its o-series models. Alongside that, he works on reinforcement learning for chain-of-thought reasoning.
-
June 2025
A move to Meta
In June 2025 Meta hires him to work on AI reasoning models.
-
Since 2025
Reasoning models at Meta
At Meta Superintelligence Labs, he works on AI reasoning models, a unit he joined in June 2025.
Sources and credits
Photo: We have no published photo of Trapit Bansal yet. Send us a free-licensed photo.
- [1710.03641] Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments
- ICLR Poster Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments
- [2009.08445] Self-Supervised Meta-Learning for Few-Shot Natural Language Classification Tasks
- Meta hires key OpenAI researcher to work on AI reasoning models | TechCrunch
- Here Is Everyone Mark Zuckerberg Has Hired So Far for Meta’s ‘Superintelligence’ Team | WIRED
- Trapit Bansal - Wikidata