Ishir Garg
AI Researcher · UC Berkeley
I'm an undergraduate at UC Berkeley studying Electrical Engineering & Computer Science and Engineering Mathematics & Statistics. I do research at Berkeley AI Research (BAIR), advised by Sergey Levine and Kurt Keutzer. Previously, I also worked with Dawn Song.
My research centers on reinforcement learning and the scalable, efficient deployment of agentic and robotic systems. Some of my previous work has centered around building principled frameworks and information-theoretic tools for generating synthetic training data and evaluating LLM agents.
My messages are always open — feel free to reach out if you're interested in talking about reinforcement learning, agentic systems, robotics, or anything in between!
News
- May 2026Interning at Google this summer! Researching self-improvement of agent harnesses.
- Apr 2026Joined PALLAS (Parallel Applications, Libraries, Languages, Algorithms, and Systems), advised by Kurt Keutzer.
- Aug 2025Joined the Robotic AI & Learning Lab (RAIL) at BAIR, advised by Sergey Levine.
Publications
Fisher-Orthogonal Projected Natural Gradient for Continual LearningUnder Review
Preprint, 2026
Research
- 2026 –Berkeley AI ResearchProf. Kurt Keutzer
- 2025 –Berkeley AI Research (RAIL)Prof. Sergey Levine
- 2025 –Berkeley AI ResearchProf. Dawn Song
Industry
- 2026 –GoogleML Engineering, Workspace GenAI
- 2026 –Narada AIFounding ML Engineer
- 2025RobloxSoftware Engineering Intern
- 2024AppleSoftware Engineering Intern
Awards
- Putnam Mathematical Competition — Top 300 (top 6% nationally)
- USA Mathematical Olympiad (USAMO) qualifier — top 250 nationally
- USA Computing Olympiad — Silver Medal