Lei M. Zhang

I am a Senior Research Scientist at Google Deepmind in Montreal. Before Deepmind, I was a Member of Technical Staff at OpenAI.

I am interested in using large-scale reinforcement learning (RL) to train highly general agents. In particular, my work have focused on using curricula with meta-RL techniques to train agents with emergent capabilities such as sim-to-real transfer P1, social learning P2, and human time-scale adaptation P3.

Please reach out if you’re interested in collaborating or chat.

Recent News:

• Apr 2023: Paper accepted to ICML 2023.
• Feb 2023: Public release of “Human-timescale Adaptation in an Open-ended Task Space”.
• Jun 2022: Public release of “Learning Robust Real-Time Cultural Transmission without Human Data”.