New blog post - The Extended PyMARL Codebase for Multi-Agent Reinforcement Learning. Explains how to install EPyMARL, run experiments, and prototype new MARL algorithms. Based on our NeurIPS Conference 2021 benchmark paper: arxiv.org/abs/2006.07869 Blog: agents.inf.ed.ac.uk/blog/epymarl/

We have open sourced github.com/nvr-avg/trajda…! It's a new, unified interface to many trajectory forecasting datasets, greatly simplifying the process of training and evaluating a forecasting model on multiple motion datasets! Boris Ivanovic NVIDIA DRIVE

New *Special Issue on Multi-Agent Systems Research in the United Kingdom* now published in AI Communications. Contains 14 contributed articles from UK labs. AI Communications The Alan Turing Institute Michael Wooldridge ➡️Editorial: content.iospress.com/articles/ai-co… ➡️Special issue: content.iospress.com/journals/ai-co…

Check out Callum Rhys Tilbury's blog to learn about discrete gradient estimators and how they can be applied to improve MADDPG's performance in discrete-action environments!

Excited to announce that we will publish a new textbook with The MIT Press @mitpress.bsky.social on Multi-Agent Reinforcement Learning in summer of 2023! With Lukas Schäfer Lukas Schäfer @EurIPS and Filippos Christianos Filippos Christianos. The book will also be available online for free. (1/4)

OUT NOW: pre-print non-final PDF of our book "Multi-Agent Reinforcement Learning: Foundations and Modern Approaches" released on official page marl-book.com. The authors (I, Filippos Christianos, Lukas Schäfer @EurIPS) will be at @aamas2023 & IEEE ICRA in London. The MIT Press @mitpress.bsky.social

Last week at @aamas2023 and IEEE ICRA #ICRA2023 was excellent. Our book marl-book.com got much attention and has been downloaded over 2500 times after one week! Huge thanks to Karl Tuyls is @karltuyls.bsky.social for featuring it in his keynote.

Our paper "Pareto Actor-Critic for Equilibrium Selection in Multi-Agent Reinforcement Learning" with Filippos Christianos Giorgos Papoudakis now accepted in Transactions on Machine Learning Research! Pareto-AC learns Pareto-optimal Equilibria in many MARL environments and reaches new sota results. arxiv.org/abs/2209.14344

It's done: finished textbook *Multi-Agent Reinforcement Learning* with Filippos Christianos Lukas Schäfer @EurIPS. Nearly 100 pages bigger than the first release in May at The AAMAS Conference & IEEE ICRA. Now going into print with The MIT Press @mitpress.bsky.social, out in late 2024. (1/2) marl-book.com

UPDATE: The copyedited version (formatting edits, index list) of the MARL book is now online! marl-book.com Next is the book cover design and then the book will go to print and stores in late 2024! Lecture slides will be released in coming weeks. The MIT Press @mitpress.bsky.social

Our Multi-Agent RL framework EPyMARL, which extends the original PyMARL framework of WhiRL, is widely used in MARL research (nearly 400 stars on Github!) and several MARL benchmarks have been built on it. We list three in this thread! Blog: agents.inf.ed.ac.uk/blog/epymarl/ (1/4)

🎉 We have an exciting update to our open-source MARL libraries for you! Our popular EPyMARL library (439⭐) with highly customisable implementations for many MARL algorithms has now become even better - welcome to EPyMARL v2.0! github.com/uoe-agents/epy… See our new blog post for

[2/3] Happy to share our new GUI Agent paper: Lightweight Neural App Control. We propose LiMAC, an architecture that balances efficiency and natural language understanding by combining a lightweight transformer with a fine-tuned VLM. Project page: arxiv.org/abs/2410.17883

Excited to share that our MARL textbook is officially launching next week with The MIT Press @mitpress.bsky.social! After plenty of writing (and rewriting), Lukas Schäfer @EurIPS, Stefano Albrecht, and I are excited to finally see it in print. Check out the thread below to learn more! 🧵👇 #RL #AI #MARL

Out NOW - Our Multi-Agent RL textbook has finally arrived in stores today! Order from The MIT Press @mitpress.bsky.social or other book stores. Just in time for Christmas! 🙂 Lecture slides (with Tex files) and algorithm code are available from the book homepage. See Lukas' post ⬇️ for more details.

The first textbook on multi-agent reinforcement learning is finally out - as I said before, a landmark for the field, the first textbook covering game-theoretic foundations with state-of-the-art deep learning! Huge congrats to its authors Stefano Albrecht , Lukas Schäfer @EurIPS and

The Multi-Agent RL book has completely sold out! 😀 Reprints are now in production by The MIT Press @mitpress.bsky.social. We have updated the book with some corrections. The book PDF + errata, slides, code and exercises are available at marl-book.com.