Harsh Mahesheka

Found a hack for bookmark feature, just repost it😂

Check out Pygame_spiel by @giogix2.bsky.social !

This is a Pygame-based library to play games from the OpenSpiel suite against AI agents. 🤩😍

github.com/giogix2/pyga...

Migrating to Bluesky feels like upgrading your codebase to that new software in the market. It's painful, but you know you will have to do it at some point.

You just witnessed the birth of Skynet

e introduce the effective horizon, a property of MDPs that controls how difficult RL is. Our analysis is mo- tivated by Greedy Over Random Policy (GORP), a simple Monte Carlo planning algorithm (left) that exhaustively ex- plores action sequences of length k and then uses m random rollouts to evaluate each leaf node. The effective horizon combines both k and m into a single measure. We prove sample complexity bounds based on the effective horizon that correlate closely with the real performance of PPO, a deep RL algorithm, on our BRIDGE dataset of 155 deterministic MDPs (right).

Kind of a broken record here but proceedings.neurips.cc/paper_files/...
is totally fascinating in that it postulates two underlying, measurable structures that you can use to assess if RL will be easy or hard in an environment

👋

Great list!! Could you add me?
I work on Multi-Agent Reinforcement Learning and it's applications in robotics. Would love to connect with community.

mesa-examples/rl at main · projectmesa/mesa-examples User and showcase agent-based models developed using Mesa - projectmesa/mesa-examples

Love to get added!!

A few of my open source projects 🤖-

1. Mesa_RL- github.com/projectmesa/...
2. Household Bot - github.com/harshmaheshe...
3. Gazebo Pkg_Create - github.com/gazebosim/gz...

This summer, I collaborated with Project Mesa to create Mesa-RL framework that seamlessly integrates multi-agent reinforcement learning into Mesa's agent-based modeling environments. 🤖🌍

Go try RL with your social agent models!

Link - github.com/projectmesa/...

#ReinforcementLearning #MultiAgent

Great List! Would love to be added and connect with the community.

I work in robot learning. Would love to be added in the list.

Hello👋, Would love to be added.

Hi, Thanks for the list.
Would love to be added to this list. I work on robot learning and its applications.

Great List!!

Would love to be added @chrispaxton.bsky.social

By chaining a VLM and LLM in a bi-level framework, we use the “chain rule” to guide reward search directly from video demos.

This work was done under guidance of Prof. Wanxin Jin and Prof. Zhorang Wang at Intelligent Robotics and Interactive Systems (IRIS) Lab, Arizona State University.

Can robots learn skills from YouTube without complex video processing? Our LLM-driven bi-level programming shows it’s possible! Check out our RL agents learning skills from their biological counterparts!💡

Preprint: arxiv.org/abs/2410.09286

#ReinforcementLearning #LLM #Robotics #AI

This was work with my great colleagues Harsh Mahesheka, Jan Ole von Hartz, Tim Welschehold and Abhinav Valada at the Technische Fakultät der Universität Freiburg, University of Freiburg.

By modularizing commands, the operator can fully focus on the task relevant end-effector motions. This even enables kinesthetic teaching of mobile manipulators in cluttered environments. We show that this allows us to rapidly learn whole-body mobile manipulation skills with less than ten minutes.

Can we operate mobile manipulators without expensive exoskeletons or tracking setups? 📜

In our latest work, Zero-Cost Whole-Body Teleoperation for Mobile Manipulation, we do exactly this.

Website and full videos: moma-teleop.cs.uni-freiburg.de
Arxiv: arxiv.org/abs/2409.150...

#AI #Robotics

I’m an Electrical Engineering student at IIT Varanasi, on a mission to teach robots human-like skills. Mimicking the most complex machine—us—is no small feat, but reinforcement learning makes it exciting (and intuitive!). 🚀🤖
#AI #ReinforcementLearning #Robotics

A starter pack of starter packs:

Robotics and AI go.bsky.app/DfAoaJ1
Computer Vision go.bsky.app/PkAKJu5
Computer Graphics Research go.bsky.app/ckQ1u9
Grumpy Machine Learners go.bsky.app/6ddpivr
Reinforcement Learning go.bsky.app/3WPHcHg