Martin Balla

Next, Martin Balla talking about challenges and opportunities for reinforcement learning in tabletop games and a research framework called PyTAG

With @developconf just under a month away. We thought we would share who's attending from iGGi this year! Looking forward to seeing folks there. More details to come

Very excited to announce that our paper Posterior Sampling for Deep Reinforcement Learning (PSDRL) has been accepted at #ICML2023! ⌨️Code: github.com/remosasso/PSDRL 📄 Paper: arxiv.org/abs/2305.00477 🧵Thread 👇

The cool thing about setting an assignment for MSc students on writing AI players for board games is that you get to run game sessions to teach the game. A bit of Sushi Go! before lunch time sounds appropriate. @GameAI_QMUL @tag_framework

We’re excited to announce that GriddlyJS will be presented at NeurIPS 2022. Design 🎨, build ⚒️, debug 🐛RL environments, record human trajectories 👲, and run policies 🤖 directly in your browser 🕸! Check it out here: griddly.ai Paper: arxiv.org/abs/2207.06105

I'm presenting our work on the #Pandemic #tabletopgame later today at @ieee_cog with @ballamist - talk recording will also premiere soon after on youtube: youtube.com/watch?v=oPhyZd… :) Always looking for #GameAI collaborations within @tag_framework, do get in touch if interested!

We’re excited to announce GriddlyJS - A Web IDE for Reinforcement Learning! Design 🎨, build ⚒️, debug 🐛RL environments, record human trajectories 👲, and run policies 🤖 directly in your browser 🕸! Check it out here: griddly.ai Paper: arxiv.org/abs/2207.06105

We highlight first TAG competition on the #boardgame #Pandemic @ieee_cog 2022 - there is still 1 month to submit #GameAI agents to cure disease in this cooperative board game! Deadline: July 29th. tabletopgames.ai/competition/co…

We are very excited to announce this: in collaboration with @CAGames, we are advertising for a PhD scholarship for @iggiphd for September 2022 start. This is a fully-funded, 4-year position for home (UK) or EU-Settled status at @GameAI_QMUL. 👇🧵 1/5

Day 1 of #GameAIHAck done, Day 2 starting soon. Plan - Try to implement a new modification for MCTS in 48h Reality - Debug and fix old code, setup everything, laugh, and play board games with @b_gum22 @ballamist @diego_pliebana #WorthIt

We at @GameAI_QMUL have been working for the last months on our framework for General Strategy Games: Stratega [🧵]

Games & the complex decision-making scenarios they offer have long helped advance AI research. In this webinar/live Q&A Feb 18 at 9 AM PT, researchers explore how Minecraft can be used to design multi-agent training tasks for reinforcement learning agents: aka.ms/AAb0twr

Done all your Christmas shopping? Thinking about maybe getting yourself a present too? How about a fully funded PhD to research games? It costs nothing except an amazing application and 4 years focusing on your research. Be part of our exciting community iggi.org.uk/apply

We've not talked much about this, but it's probably time! Presenting TAG (a tabletop games framework) at @exag20xx, ongoing project from @GameAI_QMUL with @ballamist Alexander Dockhorn, Raul Montoliu and @diego_pliebana Check it out on github! github.com/GAIGResearch/T…

Code and slides for the workshop are available on github github.com/martinballa/Ca…

Thanks to our recent funding success, applications for IGGI funded studentships starting September 2019 are again open. Join the world’s largest doctoral training centre for research in games. Deadline for applications is 23:59 GMT on March 8 2019: iggi.org.uk/apply/

Happy to see all the new @iggiphd students presenting their #gamedev projects! Lots of procedural content generation and physics. Well done everyone 😎💪