Shivakanth

Mila Quebec. NIT Trichy.

prof_pic.jpg

Hello to my small corner of the internet!

Update: Looking for full time research engineer roles starting Fall 2023!

I’m a masters student at Mila Quebec supervised by Prof. Samira Ebrahimi Kahou exploring continual reinforcement learning and computer vision for real world applications. Previously I was an research intern with John Langford at Microsoft Research NYC, working on minimal agent controllable representations in reinforcement learning.

During my masters I have had the good fortune of collaborating with researchers from Microsoft Research and Google DeepMind on projects studying sample efficiency in deep RL, from an algorithmic and evaluation perspective.

I completed my undergrad in Intrumentation and Control Engineering from NIT Trichy. Feel free to reach out to talk about reinforcement learning or control theory.

You can find my CV here.

news

Oct 11, 2023 Our paper on “Bridging the Gap Between Offline and Online Reinforcement Learning Evaluation Methodologies” has been accepted at TMLR 2023! 📣
Sep 28, 2023 Our work on “Prioritizing Samples in Reinforcement Learning with Reducible Loss” has been accepted at NeurIPS 2023! 📣
Jun 12, 2023 Started as a research intern at Microsoft Research NYC with John Langford.
Sep 14, 2022 Two papers accepted to NeurIPS 2022 workshops! Come visit us at the Deep RL and Offline RL workshops. ✈
Sep 14, 2022 Our work on “Learning Robust Dynamics through Variational Sparse Gating” has been accepted at NeurIPS 2022! 📣

selected publications

  1. Under Review
    PcLast : Discovering Plannable Continuous Latent States
    Anurag Koul, Shivakanth Sujit, Shaoru Chen, and 8 more authors
    Under Review, ICLR, 2024
  2. NeurIPS
    Prioritizing Samples in Reinforcement Learning with Reducible Loss
    Shivakanth Sujit, Somjit Nath, Pedro HM Braga, and 1 more author
    NeurIPS, 2023
    arXiv Code Website
  3. TMLR
    Bridging the Gap Between Offline and Online Reinforcement Learning Evaluation Methodologies
    Shivakanth Sujit, Pedro HM Braga, Jorg Bornschein, and 1 more author
    Transactions on Machine Learning Research (TMLR), 2023
    arXiv Code Website
  4. NeurIPS
    Learning Robust Dynamics through Variational Sparse Gating
    Arnav Kumar Jain, Shivakanth Sujit, Shruti Joshi, and 3 more authors
    NeurIPS, Oct 2022
    arXiv Code