Talks

Research Presentations

• Provably Efficient CVaR RL in Low-rank MDPs
  [poster]

• Local Optimization Achieves Global Optimality in Multi-Agent Reinforcement Learning
  [slides] [poster]

• Blessing of Class Diversity in Pre-training
  [slides] [poster]

• Optimizing the Performative Risk under Weak Convexity Assumptions
  [slides] [poster]

• Provably Efficient Policy Optimization for Two-Player Zero-Sum Markov Games
  [slides] [poster]

Other Works

• Enhancing Network Pruning Method Evaluation and Integration
  [report] [poster]

• Can Learned Models Replace Hash Functions?
  [slides]

• Reinforcement Learning From Human Feedback
  [slides]

• Training Diffusion Models with Reinforcement Learning
  [slides]

• Risk Sensitive Reinforcement Learning with Low-rank MDPs
  [slides]

• Eigenface++: Face Recognition with Deep PCA
  [report] [slides]

• What Can Transformers Learn In-Context?
  [slides]