Ryan Lindeborg

I’m a Research Affiliate at MIT CSAIL, where I work with Will Shen, Nishanth Kumar, Leslie Kaelbling, and Tomás Lozano-Pérez, and a master’s student at USC, where I work with Abrar Anwar, Jesse Zhang, and Jesse Thomason. My research interests are in robot learning and aligning robot behavior with human intent.

Previously, I developed large-scale robotic systems at Amazon Robotics and graduated from Harvard.

Research

  1. Anticipating Unintended Robot Behaviors with Side Effect Critics
    Ryan Lindeborg*, Richard Jiankun Peng*, Ayush Jain, Jesse Zhang, Abrar Anwar, and Jesse Thomason
    Spotlight Presentation, Rethinking Safety Workshop, RSS 2026
    We study side effect prediction for robotic manipulation, aiming to anticipate unintended changes before they occur, so the robot can stop, resample, or defer to a human. We train a side effect critic, initialized from a pre-trained vision-language action (VLA) model, that predicts whether given action chunks will cause a side effect before the actions are executed.
  2. TiPToP: A Modular Open-Vocabulary Robot Manipulation System That Plans
    William Shen*, Nishanth Kumar*, Sahit Chintalapudi, Ryan Lindeborg, Jie Wang, Christopher Watson, Edward S. Hu, Jing Cao, Dinesh Jayaraman, Leslie Pack Kaelbling, and Tomás Lozano-Pérez
    FM4RoboPlan Workshop, RSS 2026
    TiPToP is a modular manipulation system that integrates pre-trained foundation models with a GPU-accelerated Task and Motion Planner to solve tasks directly from RGB images and natural language.