I'm JB (John Banister) Lanier. I specialize in Deep Reinforcement Learning for multi-agent systems and robotics. Much of my work in the past few years has focused on efficiently training strong agents for two-player competitive games. Currently, I'm working on simulation-to-real transfer for robotic control, and I'm creating new model-based RL methods that are better at adapting to real environments.
I want to make systems that generate plannable world models for brand-new environments and reliable agents that can react to the unexpected.
Feel free to reach out to me if you're interested in chatting or collaborating!
____________________________