I'll be presenting an overview of learning Mean Field Games and Mean Field Control with Reinforcement Learning next Tuesday 👀

Today 5pm CET!

Very happy to co-organize the Gamification and Multiagent Solution workshop at ICLR 2025 2022! We are accepting papers, more infos on our website. --> gamificationmas.com

Submissions are open 🤗 Submission Deadline: February 26, 2022 (AoE). Acceptance Notification: March 26, 2022. --> Workshop: April 29, 2022.

This work was led by Matthieu Geist in a collaboration between Google AI, @Deepmind, and Inria. It is joint work between Matthieu Geist, Julien Pérolat, Mathieu Laurière, Romuald Elie, Sarah Perrin, myself, Rémi Munos, and Olivier Pietquin.

🏅💸💰 AWARD ALERT 💰💸🏅 The Cooperative AI Foundation (CAIF) kindly offered to give 2 awards of 500$ each for “Best Cooperative AI paper” and “Best Cooperative AI poster”! We are very grateful and waiting for your submissions 😎 cooperativeai.com/foundation

Today at 6PM CET, Sarah Perrin is going to present her work on learning master policies that leverage generalisation properties of mean-field games. A master policy leads to a Nash equilibrium irrespective of the initial distribution. Visit us at R3B for more! 🙏 #AAAI2022

The list of accepted papers is on our website: gamificationmas.com/accepted-papers Check this out!

Happening now!

First paper accepted ICML Conference! And this means... first in-person conference 🥳 --> "Scalable Deep Reinforcement Learning Algorithms for Mean Field Games" arxiv.org/abs/2203.11973

Today, our PhD student Sarah Perrin is going to give an invited talk at #EWRL2022 in Milan. At 9 AM, she will talk about "Scaling up MARL with MFGs and vice versa". Stay tuned to know more of Scool's presence at #EWRL2022. 😃🙏

Welcome to #EWRL2022 day 2! Today's invited speakers are Sarah Perrin, Niao He, Ann Nowe and Richard Sutton.

Introducing Gemini 1.0, our most capable and general AI model yet. Built natively to be multimodal, it’s the first step in our Gemini-era of models. Gemini is optimized in three sizes - Ultra, Pro, and Nano Gemini Ultra’s performance exceeds current state-of-the-art results on

Happy to introduce our paper MusicRL, the first music generation system finetuned with human preferences. Paper link: arxiv.org/abs/2402.04229

Introducing Gemma - a family of lightweight, state-of-the-art open models for their class, built from the same research & technology used to create the Gemini models. Blog post: blog.google/technology/dev… Tech report: goo.gle/GemmaReport This thread explores some of the

Gemma-2-27B-IT is "now the best open model!" 🚀🚀🚀

Check out our new RLHF algorithm ⬇️

Are small models still undertrained? We are releasing a 2B model that beats GPT-3.5. The crazy part is that it was distill on only 2T tokens from a small model. Distillation is the future of LLMs with the growing availability of large and efficient open models!