Jagdeep Singh Bhatia

Hey there! I'm Jagdeep, a sophomore at MIT studying computer science.

I'm interested in thinking about some of the more nuanced problems in AI (interpretability, robustness, generalizability) and applying those ideas to designing intelligent embodied systems. I'm really proud of EvoGym, a benchmark for design and control co-optimization for voxel-based soft robots which I built with the Computational Design and Fabrication Group at MIT. I spent this winter designing debiasing algorithms at Themis AI, and this summer, I'll be working on similar challenges at Scale. My work has been featured in Scientific American, Wired, Forbes, IEEE Spectrum, and MIT News.

In my free time I love running, playing card games, and making bad puns.


Jun 2022

   Black-box attacks and vulnerabilities of Apple’s NEURALHASH exposed and published in ICML, ML4Cyber Workshop.

Jan 2022

   Open sourced docs & codebase for Evolution Gym – a benchmark for design and control co-optimization for voxel-based soft robots.

Dec 2021

   Accepted position as Machine Learning Researcher @ Scale AI for Summer ‘22.

Dec 2021

   Evolution Gym is featured in Scientific American, Wired, MIT News, and IEEE Spectrum!

Sep 2021

   Research on co-design, evolutionary robots, and reinforcement learning published in NeurIPS.

Dec 2020

   Research on simple and fast interactive machine learning algorithms published in JMLR.

Jul 2020

   Won $175K and 2nd Place in the Regeneron Science Talent Search, a research and science competition for high school seniors.



  1. Exploiting and Defending Against the Approximate Linearity of Apple’s NeuralHash.
    Jagdeep Singh Bhatia, and Kevin Meng.

    In International Conference on Machine Learning 2022.


  1. Simple and Fast Algorithms for Interactive Machine Learning with Random Counter-examples.
    Jagdeep Singh Bhatia.

    In Journal of Machine Learning Research 2021.

  2. Evolution Gym: A Large-scale Benchmark for Evolving Soft Robots.
    Jagdeep Singh Bhatia, Holly Jackson, Yunsheng Tian, Jie Xu, and Wojciech Matusik.

    In Advances in Neural Information Processing Systems 2021.