Matej Jusup @matejjusup.bsky.social · 05/08/20251/N I’ve long believed that board games should play a bigger role in AI evaluation. They naturally test strategic reasoning, long-term planning, adaptation—and they can’t be solved by brute force or memorization. Game Arena is transparent, replayable, and tests actual behavioral intelligence. 131
Matej Jusup @matejjusup.bsky.social · 23/05/2025A year after our trip to AAMAS in New Zealand, @sharky6000.bsky.social came back for more! I should have planned my year not to miss @aamasconf.bsky.social… Big congrats and keep up amazing work! 🎉👏 130
Matej Jusup @matejjusup.bsky.social · 20/05/2025Looking forward to speaking at the ML Pub Club on June 3rd! I'll discuss how, during my time at DeepMind, we taught LLMs to play chess at a GM level and the broader implications for strategic AI. If you're in Zagreb, join us at Mažuranićev trg 13 at 6 PM! More info & RSVP: lu.ma/erjji5itlu.maML Pub Club #22: Superhuman Planning with LLMs · LumaWhat happens when a chess champion meets cutting-edge AI? Join us for an evening with Matej Jusup, as he unpacks how large language models (LLMs) can go from… 030
Matej Jusup @matejjusup.bsky.social · 01/05/2025A paper from my time at Google was accepted for a spotlight presentation at ICML! In “Mastering Board Games by External and Internal Planning with Language Models”, we show how language models can achieve grandmaster-level play using a search budget on par with humans. arxiv.org/abs/2412.12119arxiv.orgMastering Board Games by External and Internal Planning with Language ModelsAdvancing planning and reasoning capabilities of Large Language Models (LLMs) is one of the key prerequisites towards unlocking their potential for performing reliably in complex and impactful domains... 0203
Reposted by Matej JusupMarc Lanctot @sharky6000.bsky.social · 28/04/2025Hive (and all of its expansions) has been added to OpenSpiel! 🎉🤩🐝🐜🕷️🐞🦟🪲 From Gen42: "Hive is an award-winning board game with a difference. There is no board. The pieces are added to the playing area thus creating the board. As more and more pieces are added the game becomes a fight to ... 🧵1/5 1143
Reposted by Matej JusupCsaba Szepesvari @skiandsolve.bsky.social · 06/03/2025www.youtube.com/watch?v=9_Pe... An interview with Rich. The humility of Rich is truly inspiring: "There are no authorities in science". I wish people would listen and live by this.youtube.comTURING AWARD WINNER Richard S. Sutton in Conversation with Cam Linke | No Authorities in ScienceYouTube video by Amii 24013
Reposted by Matej JusupDaphne Cornelisse @daphne-cornelisse.bsky.social · 28/02/2025Sim agents are key for developing autonomous systems for safety-critical systems, like self-driving cars. We're open-sourcing sim agents that achieve a 99.8% success rate with < 0.8% failures on the Waymo Dataset. These agents are built through scaling self-play. 3335
Reposted by Matej JusupAndreas Krause @arkrause.bsky.social · 17/02/2025We've released our lecture notes for the course Probabilistic AI at ETH Zurich, covering uncertainty in ML and its importance for sequential decision making. Thanks a lot to @jonhue.bsky.social for his amazing effort and to everyone who contributed! We hope this resource is useful to you! 16410
Matej Jusup @matejjusup.bsky.social · 14/02/2025LLMs Mastering Board Games: ZurichNLP Meetup - Feb 20th! Excited to share insights from my student research at Google DeepMind at the upcoming ZurichNLP meetup! I'll present how we achieved high-level play in board games using LLMs with a search budget comparable to human chess grandmasters. 2153
Reposted by Matej JusupEugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 09/02/2025I've been talking about writing this paper to anyone who would listen since 2020. I bombed a bunch of job talks trying to convince companies to work on this. It's so nice to finally just be able to say, yes, self-play RL in a diverse world gives you immense capabilities arxiv.org/abs/2502.03349arxiv.orgRobust Autonomy Emerges from Self-PlaySelf-play has powered breakthroughs in two-player and multi-player games. Here we show that self-play is a surprisingly effective strategy in another domain. We show that robust and naturalistic drivi... 3926
Reposted by Matej JusupNikola Zubić / Никола Зубић @nikolazubic.bsky.social · 04/02/2025I am more than happy that @quantamagazine.bsky.social , which I have been reading since the first year of my Bachelor's degree, cited us: www.quantamagazine.org/chatbot-soft... More news about this work and 2nd version is coming soon! #machinelearning #deeplearning #cs #computerscience #tcs 041
Reposted by Matej JusupCathy Wu @cathywu.bsky.social · 29/01/2025Pet peeve: Calling something that’s not open source… open source. Open weight != open source 1303
Reposted by Matej JusupBrent Toderian @brenttoderian.bsky.social · 25/01/2025A typical European car is parked 92% of the time. It spends 1/5th of its driving time looking for parking. Its 5 seats only move 1.5 people. 86% of its fuel never reaches the wheels, and most of the energy that does, moves the car, not the people. Sound efficient? HT @ellenmacarthurfdn.bsky.social 301088404
Matej Jusup @matejjusup.bsky.social · 26/01/2025An interesting idea that’s worth keeping an eye on! 000
Reposted by Matej JusupJeff Dean @jeffdean.bsky.social · 24/01/2025Demis Hassabis, James Manyika, and I wrote up an overview of the AI research work & advances across Google in 2024 (Gemini, NotebookLM, robotics, ML for science, & advances in responsible AI+more). 🎊 Given it a read or paste it into NotebookLM to listen, if you prefer! blog.google/technology/a...blog.google2024: A year of extraordinary progress and advancement in AIAs we move into 2025, we’re looking back at the astonishing progress in AI in 2024. 212423
Reposted by Matej JusupMarc Lanctot @sharky6000.bsky.social · 20/01/2025Check out the 16th Workshop on Optimization and Learning in Multiagent Systems (OptLearnMAS-25) at #AAMAS 2025! Topics: distributed opt., coalition formation, opt. under uncertainty, winner determination algs in auctions and procurements, algs to compute equilibria in games. optlearnmas.github.io 0112
Reposted by Matej JusupMarc Lanctot @sharky6000.bsky.social · 17/01/2025In December, I posted about our new paper on mastering board games using internal + external planning. 👇 Here's a talk now on Youtube about it given by my awesome colleague John Schultz! www.youtube.com/watch?v=JyxE...youtube.comJohn Schultz, DeepMind, Mastering Board Games by External and Internal Planning with Language ModelsYouTube video by AI4All 13511
Matej Jusup @matejjusup.bsky.social · 17/01/2025John's talk is now available online! www.youtube.com/watch?v=JyxE...youtube.comJohn Schultz, DeepMind, Mastering Board Games by External and Internal Planning with Language ModelsYouTube video by AI4All 0132
Matej Jusup @matejjusup.bsky.social · 13/01/2025Join John's talk to get insights on our paper on mastering board games with language models! 061
Reposted by Matej JusupMarc Lanctot @sharky6000.bsky.social · 10/01/2025Just a reminder that the AAMAS Doctoral Consortium deadline is next Friday! Please consider submitting to this great venue or telling your students about it. 👇 083
Matej Jusup @matejjusup.bsky.social · 03/01/2025After 15 years away from competitive chess, I forgot how much thrill and excitement the game gives! ♟️ I decided to attend a tournament with five grandmasters and numerous international, fide, and candidate masters. @lichess.org broadcast: lichess.org/broadcast/29... 130
Reposted by Matej JusupCsaba Szepesvari @skiandsolve.bsky.social · 19/12/2024If you are into ML theory (RL or not) with a proven track record, and you are interested in an industry research position, PM me. Feel free to spread the word. 27531
Matej Jusup @matejjusup.bsky.social · 18/12/2024After a slight delay, it is now also out on arXiv: arxiv.org/abs/2412.12119 081
Matej Jusup @matejjusup.bsky.social · 16/12/2024I will remember @neuripsconf.bsky.social 2024 as a defining moment in my career. Grateful to mentors & colleagues at @ethzurich.bsky.social and @deepmind.google.web.brid.gy for making it possible. Meeting enthusiastic researchers and having insightful, constructive discussions was truly inspiring! 050
Matej Jusup @matejjusup.bsky.social · 12/12/2024Don’t miss this talk by @sharky6000.bsky.social if you want to turn LLMs into interactive, gamified chatbots! 020
Matej Jusup @matejjusup.bsky.social · 11/12/2024I was honored to present our demo on “Mastering Chess With Language Models” at the @GoogleDeepMind booth at @NeurIPSConf right before an inspiring talk by Jeff Dean Thanks to everyone who dropped by and made it a memorable session! ♟️ 2162
Reposted by Matej JusupEric Malmi @ericmalmi.bsky.social · 11/12/2024if you're at #NeurIPS2024, want to learn how to make LLMs really good at chess, and see a live demo, come and visit the Google DeepMind booth tomorrow at 9:30 am! 0182
Matej Jusup @matejjusup.bsky.social · 10/12/2024I couldn’t ask for a better @neuripsconf.bsky.social welcome! Spotted a lovely family playing chess and had the joy of facing a brilliant 6-year-old who just learned the game 3 days ago. Her enthusiasm was truly inspiring! 140
Matej Jusup @matejjusup.bsky.social · 07/12/2024I will be showing a “Mastering Chess With Language Models” demo at @googledeepind.bsky.social booth at @neuripsconf.bsky.social with @ericmalmi.bsky.social @nenadtomasev.bsky.social @anianruoss.bsky.social Don't hesitate to reach out if interested or want to chat! deepmind.google/discover/eve...deepmind.googleNeurIPS 2024NeurIPS 2024 1252
Matej Jusup @matejjusup.bsky.social · 06/12/2024𝗚𝗼𝗼𝗱 𝗲𝗻𝗼𝘂𝗴𝗵 𝗮𝘁 𝗰𝗵𝗲𝘀𝘀 𝘁𝗼 𝗼𝘂𝘁𝘀𝗺𝗮𝗿𝘁 𝗮 𝗹𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝗺𝗼𝗱𝗲𝗹? ♟️ 𝗚𝗶𝘃𝗲 𝗶𝘁 𝗮 𝘀𝗵𝗼𝘁! 👉 gemini.google.com/gem/chess-ch... Share your victories (and brags) in the comments! 2101
Matej Jusup @matejjusup.bsky.social · 05/12/2024"𝗖𝗮𝗻 𝗰𝗵𝗲𝘀𝘀 𝗮𝗻𝗱 𝗼𝘁𝗵𝗲𝗿 𝗯𝗼𝗮𝗿𝗱 𝗴𝗮𝗺𝗲𝘀 𝘀𝘁𝗶𝗹𝗹 𝗮𝗱𝗱 𝘃𝗮𝗹𝘂𝗲 𝘁𝗼 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵?” I was fortunate to join the board games team at Google as a student researcher, where we proved that the answer is a resounding YES! ♟️🤯 deepmind.google/research/pub... 1266
Reposted by Matej JusupMarc Lanctot @sharky6000.bsky.social · 05/12/2024Super happy to reveal our new paper! 🎉🙌♟️ We trained a model to play four games, and the performance in each increases by "external search" (MCTS using a learned world model) and "internal search" where the model outputs the whole plan on its own! 413818
Matej Jusup @matejjusup.bsky.social · 05/12/2024The project I was quite honored to work on! ♟️ As a chess enthusiast, I was especially thrilled that an LLM equipped with an efficient MCTS reaches grandmaster-level strength with a search budget comparable to human players! 🤯 deepmind.google/research/pub... 161