PipelineRL: RL training speed through the roofline
Talk, Poolside, Virtual
Talks and presentations
LLMs can learn self-restraint through iterative self-reflection
Talk, ServiceNow Research, Montreal, Qc
Probabilistic Planning with Sequential Monte Carlo methods
Conference proceedings talk, Workshop on Probabilistic Reinforcement Learning and Structured Control @ Neurips 2018, Montreal, Qc
AI safety and reward misspecification
Talk, Brookfield Institute for Innovation + Entrepreneurship, Toronto, Ontario
