Scholar
Stone Tao
Google Scholar ID: GAMO0EwAAAAJ
University of California - San Diego
Citations & Impact
Citations
607
H-index
8
i10-index
6
Publications
14
Co-authors
23
list available
Publications
Resume (English only)
Academic Achievements
- Paper 'Staggered Environment Resets Improve Massively Parallel On-Policy Reinforcement Learning' accepted to NeurIPS 2025.
- ManiSkill3 accepted to RSS 2025 and selected for an oral presentation at the ICLR Robot Learning Workshop 2025.
- Two papers accepted to ICLR 2025.
- Contributed to ManiSkill-HAB and Policy Decorator, both published at ICLR 2025.
- Lux AI Challenge Season 2 accepted as a competition at NeurIPS 2023.
- Awarded the NSF Graduate Research Fellowship in March 2023.
- Led the release of ManiSkill3 Beta: a state-of-the-art, fully open-source, GPU-parallelized robotics simulation platform enabling high-speed visual RL even on Google Colab.
Background
- Third-year PhD student at UC San Diego advised by Professor Hao Su.
- Currently a research intern at NVIDIA Research.
- Research interests center on advancing embodied AI and robot foundation models using compute-scalable synthetic data such as simulation or world models.
- Focuses on machine learning tools for robotics, especially reinforcement learning (RL), advocating a 'simulation integrated machine learning' paradigm.
- Passionate about building and running high-quality open-source AI competitions for education and research.
