2025: Paper 'Learning Manipulation Skills through Robot Chain-of-Thought with Sparse Failure Guidance' accepted by IROS 2025.
2025: Paper 'Video2Policy: Scaling up Manipulation Tasks in Simulation through Internet Videos' published on ArXiv.
2024: Paper 'Reinforcement Learning with Foundation Priors: Let the Embodied Agent Efficiently Learn on Its Own' accepted as Oral at CoRL 2024 (top 4%).
2024: Paper 'EfficientZero V2: Mastering Discrete and Continuous Control with Limited Data' accepted as Spotlight at ICML 2024 (top 3.5%).
2024: Paper 'Seer: Language Instructed Video Prediction with Latent Diffusion Models' accepted at ICLR 2024.
2023: Paper 'Become a Proficient Player through Watching Pure Videos' accepted at ICLR 2023 (equal contribution).
2023: Paper 'SpeedyZero: Mastering Atari with Limited Data and Time' accepted at ICLR 2023.
2022: Two papers accepted at NeurIPS 2022 on MCTS acceleration and sample-efficient imitation learning.
2021: Paper 'Mastering Atari Games with Limited Data' accepted at NeurIPS 2021.
2020 & 2019: Published papers at AAAI on cross-domain sentiment classification and domain adaptation; the 2019 paper was an Oral presentation.