Try 💎 DIAMOND’s Counter Strike world model directly in your browser! → next.journee.ai/xyz-diamond ← How long can you stay in distribution? Can you beat Stéphane Liem Nguyen’s 1000 frames? Adam Jelley and I are at NeurIPS! Poster #6306, Friday 11am-2pm, West Ballroom

You can now play fully generative games on an interactive video model, and we’ve collected together them together at worldarcade.gg. You can now play Mario Kart 64, Minecraft, CS:GO and many other games on your own device. All open source.

1/ Does scaling pre-training 📈 improve embodied AI 🤖as effectively and predictably as it is does LLMs? arxiv.org/abs/2411.04434 Yes! 👍 Our paper studies behavior cloning & world modeling tasks in the infinite data regime ♾️on human data🎮

Awesome to see this released and out in the world now. Huge congrats to the entire team! It was super cool to see the vision come to life as an intern with everything coming together 👏

Today in Nature: our research on world and human action models (WHAM) - generative ai models of video games, aimed towards supporting game creatives in gameplay ideation : nature.com/articles/s4158… - huge congrats to everyone who made this happen, I couldn't be more proud 🥳

Introducing M³: A 𝗠odular 𝗪orld 𝗠odel over streams of tokens for sample-efficient RL 🌍🤖 M³ achieves state-of-the-art performance for planning-free world models on Atari-100K 🕹️, DMC 🦾, and Craftax-1M! 🚀 🧵1/8

One week left to submit your contributions for our workshop RL_Conference... gotta go fast! ⏰🎮 We welcome submissions on video games for benchmarking RL and RL for video game development. ⌛Paper Submission Deadline: May 30th Find out more here: sites.google.com/view/rlvg-work…

We’re excited to announce our first speaker: James MacGlashan from Sony AI! 🚀 James led the development of GT Sophy, a superhuman RL racing agent in Gran Turismo. 🏎️ Join his talk to learn about how GT Sophy was used as an opponent in GT7 and what’s next for RL in gaming. 🎮

We’re excited to announce our next speaker: Roberta Raileanu (Roberta Raileanu) from Google DeepMind! Roberta will discuss NetHack: A Grand Challenge for RL and LLM Agents Alike. ⚔️ Join us on August 5th to learn how to develop agents capable of tackling open-ended environments!

Thank you to our headline sponsors Medal for their kind support to make this workshop happen - go check them out at medal.tv to capture and share your PC gaming highlights! 🎮🕹️

We’re excited to announce our next speaker: Pablo Samuel Castro from Google DeepMind! Pablo Samuel Castro focuses on fundamental RL research. He will make the case that the Atari benchmark remains relevant for advancing our scientific understanding of RL.🕹️ See if you agree RL_Conference!

We’re excited to announce our final keynote speaker: Michael Bowling (University of Alberta)! 🎲🃏 Michael will explain what led to him to pioneer the use of old video games to test RL agents. Join us RL_Conference to find out more about the story behind the Atari Benchmark!

Missed our workshop at RL_Conference? Watch all keynote presentations and the panel discussion on YouTube: youtube.com/playlist?list=… Big thanks to our awesome speakers: Julian Togelius, James MacGlashan, Roberta Raileanu, Michael Bowling, and Pablo Samuel Castro!

We’re starting a company to build the next frontier of embodied intelligence - General Intuition We believe the time has come to move from words to worlds, and we’re developing foundation models capable of acting in any environment. Starting with video games!

On a random Saturday last October, Pim de Witte reached out about a wild, large and incredibly diverse dataset of gaming clips. I was skeptical - such a dataset felt like a secret archaeological treasure. We dug. It’s real.

From the first meeting with the one and only Pim de Witte last October, to finishing my PhD and co-founding General Intuition together with Eloi Alonso and Vincent Micheli earlier this year, the last year has been quite the journey to say the least!