Profile photo

Akarsh Kumar

[Twitter] [Google Scholar] [GitHub]

Contact akarshkumar0101(at)gmail(dot)com

I'm a Ph.D. student at MIT CSAIL working with Phillip Isola.
I'm also a research intern at Sakana AI working with Yujin Tang and David Ha.
My research is supported by the NSF GRFP!

My research interests include:

  • Applying principles from natural evolution and artificial life to create better AI systems.
  • Open-ended processes which keep creating interesting artifacts indefinitely.
  • Emergent intelligence from scratch, without the internet, like how natural evolution created us.
In practice, I'm espcially interested in meta-learning, RL, automatic environment generation, and multi-agent self-play.

Research

Automating the Search for Artificial Life with Foundation Models

Akarsh Kumar, Chris Lu, Louis Kirsch, Yujin Tang, Kenneth O. Stanley, Phillip Isola, David Ha

arXiv 2024

TLDR: A new ALife paradigm using CLIP to search for (1) target, (2) open-ended, and (3) diverse simulations.
Helps to understand "life as it could be" in arbitrary substrates.

[Paper (web)] [Paper (pdf)] [Blog] [Code] [Tweet]

Learning In-Context Decision Making with Synthetic MDPs

Akarsh Kumar, Chris Lu, Louis Kirsch, Phillip Isola

AutoRL @ ICML 2024

TLDR: In-context RL agents trained on only synthetic MDPs generalize to real MDPs.

[Paper]

Effective Mutation Rate Adaptation through Group Elite Selection

Akarsh Kumar, Bo Liu, Risto Miikkulainen, Peter Stone

GECCO 2022

TLDR: Why genetic algorithms fail to self-adapt their mutation rate and how group selection fixes it.

[Paper] [Code]

Physically Plausible Pose Refinement using Fully Differentiable Forces

Akarsh Kumar, Aditya R. Vaidya, Alexander G. Huth

EPIC @ CVPR 2021

TLDR: Accurately refining pose estimations by differentiably modeling the physics of the scene.

[Paper] [Presentation]