Portrait of Rosario Scalise

Rosario Scalise

PhD Candidate (ABD)

Machine Learning & Robotics, University of Washington

Rosario is a PhD candidate at the University of Washington in ImageMachine Learning & Robotics. His research, in the ImageRobot Learning Lab, applies machine learning and optimization to continual learning, reinforcement learning, large-scale physics-based simulation, and the science of transferring learned behaviors to physical systems.

News

Selected works +16 more

  1. Preprint 2026
    A Balanced Data Diet: Addressing the Bottleneck in Mega-Scale RL for Robot Control (SGS) tl;drBalancing the data mix, not just scaling, enables effective learning with >1M+ parallel rollouts. O. Zhang*NVIDIA, M. G. Castro*, P. Yin*, A. Gupta, R. Scalise*†, & B. Boots†. project
    Manipulation · 1 / 2
  2. ICRA 2025
    Using non-expert data to robustify imitation learning via offline reinforcement learning (RISE) tl;drOffline RL on cheap non-expert data makes imitation policies robust to mistakes experts never demonstrate. K. Huang*, R. Scalise*, C. Winston, A. Agrawal, Y. Zhang, R. Baijal, M. Grotz, B. Boots,
    B. BurchfielToyota Research Institute, M. Itkina, P. Shah, & A. Gupta. arxiv
    ▶ media · 16:9
  3. ICLR 2025 @ CoRL WS
    Model predictive adversarial imitation learning for planning from observation (MPAIL) tl;drPairing model-predictive control with adversarial imitation lets a robot plan from observation alone, no action labels. T. Han, Y. Bao, B. Mehta, G. Guo, A. Vishwakarma, E. Kang, S. Jung, J. Zhou, B. Xu,
    R. Scalise, & B. Boots. arxiv · follow-up · Best Paper Finalist
    ▶ media · 16:9
  4. CoRL 2025
    Demonstrating Wheeled Lab: Modern Sim2Real for Low-cost, Open-source Wheeled Robotics T. Han, P. Shah, S. Rajagopaal, Y. Bao, S. K. Jung, S. S. Talia, G. Guo, B. Xu, B. Mehta, E. Romig,
    R. Scalise, & B. Boots. arxiv · projectNVIDIAImage LycheeAI
    ▶ media · 16:9
  5. ICRA 2024
    Toward Self-Righting and Recovery in the Wild: Challenges and Benchmarks R. Scalise, E. Caglar, B. Boots, & C.C. Kessens.
    ▶ media · 16:9
  6. ICRA 2024
    Open X-Embodiment: Robotic learning datasets and RT-X models …, R. Scalise, …Google DeepMind. · Best Paper AwardIEEE SpectrumImage TechCrunch Google DeepMind AI News Synced
    ▶ media · 16:9
  7. RSS 2024
    DROID: A large-scale in-the-wild robot manipulation dataset A. Khazatsky*, K. Pertsch*, …, R. Scalise, …, C. FinnStanford. project
  8. CoRL 2024 Workshop
    Parental Guidance: Efficient Lifelong Learning through Evolutionary Distillation (PG-1) O. Zhang, Q. Peng, R. Scalise†, & B. Boots†. arxiv · Best Paper Award
    ▶ media · 16:9
  9. ICRA 2023
    GuILD: Guided Incremental Local Densification for Accelerated Sampling-based Motion Planning R. Scalise, A. MandalikaAurora, B. Hou, S. Choudhury, & S.S. Srinivasa. ieee
    ▶ media · 16:9
  10. IJRR 2018
    Natural language instructions for human-robot collaborative manipulation R. Scalise, S. Li, H. AdmoniCarnegie Mellon, S. Rosenthal, & S.S. Srinivasa. github
    ▶ media · 16:9

Mentoring

Elsewhere

© 2026 Rosario Scalise · terms