Kunal Jha

@kjha02

CS PhD student @UW, Student Researcher @Google prev. CSxPhilosophy @Dartmouth

kjha02.github.io

Joined March 2024

Replies Reposts Media

Pinned
Kunal Jha
@kjha02
Sep 11
Can self-interested, self-improving, self-replicating agents learn to cooperate? Our new paper, Tapes Together Strong, shows they can: when social behavior, computation, and reproduction share one energy budget, cooperation evolves from scratch. arxiv.org/abs/2609.10817 🧵
GIF
32
Kunal Jha
@kjha02
19h
Excited to present this work @NeurIPSConf this December! Thanks again to my collaborators @fcicala @blaiseaguera @tyrell_turing @natashajaques @maxhkw @eyvindn !!!
Kunal Jha
@kjha02
Sep 11
Can self-interested, self-improving, self-replicating agents learn to cooperate? Our new paper, Tapes Together Strong, shows they can: when social behavior, computation, and reproduction share one energy budget, cooperation evolves from scratch. arxiv.org/abs/2609.10817 🧵
GIF
Kunal Jha
@kjha02
Jul 21
In Rio this week for CogSci! Reach out if you’d like to chat about (multi-) agent systems, RL, evolution, game theory or anything else!
Kunal Jha
@kjha02
Apr 8
Really excited to have the opportunity to give a talk on this work @cogsci_soc !!! Last year was so much fun can’t wait to go back to Rio in July 🇧🇷 Literally can’t thank my collaborators enough @aydan_huang265 @EricYe29011995 @natashajaques @maxhkw 🙏
1
Kunal Jha
@kjha02
Jun 8
I made an interactive blog post on tabular RL, inspired by some of @RichardSSutton 's earlier demos. It started as a brainstorming tool, but I hope it’s educational! Edit the update rules and watch agents learn in real time:
kjha02.github.io
Interactive Tabular RL in Four Rooms
CS PhD @ UW, CS & Philosophy @ Dartmouth College
Kunal Jha
@kjha02
Jun 2
More task variety isn't always a silver bullet for RL! We found that while diversity drives zero-shot transfer, it actually bottlenecks continual learning. Really excited to see how this tradeoff shapes the way we design future agentic systems. Huge congrats to the team!🚀
Wilka Carvalho
@cogscikid
Jun 2
Task diversity is supposedly key to generalization in RL. But what does it do to continual RL, where agents face one new task distribution after another? We find that past a point, more diversity actually inhibits continual reinforcement learning 🧵
00:00