Search Results

< Back GAIG Meetup The recent Game AI Meetup took place on 01 March 2023. Talks and presentation included: Jakob Foerster (University of Oxford, UK): Opponent-Shaping and Interference in General-Sum Games Original talk abstract: In general-sum games, the interaction of self-interested learning agents commonly leads to collectively worst-case outcomes, such as defect-defect in the iterated prisoner's dilemma (IPD). To overcome this, some methods, such as Learning with Opponent-Learning Awareness (LOLA), shape their opponents' learning process. However, these methods are myopic since only a small number of steps can be anticipated, are asymmetric since they treat other agents as naive learners, and require the use of higher-order derivatives, which are calculated through white-box access to an opponent's differentiable learning algorithm. In this talk I will first introduce Model-Free Opponent Shaping (M-FOS), which overcomes all of these limitations. M-FOS learns in a meta-game in which each meta-step is an episode of the underlying (``inner'') game. The meta-state consists of the inner policies, and the meta-policy produces a new inner policy to be used in the next episode. M-FOS then uses generic model-free optimisation methods to learn meta-policies that accomplish long-horizon opponent shaping. I will finish off the talk with our recent results for adversarial (or cooperative) cheap-talk: How can agents interfere with (or support) the learning process of other agents without being able to act in the environment? Vanessa Volz ( modl.ai ): Establishing Trust in AI-based Tools for Game Development Original talk abstract: AI-based tools to support the game development process have long been a topic in Game AI research, with popular publications in testing, churn prediction, asset, level and even game generation. However, the adaptation of these techniques from the games industry has been hesitant at best: The small-scale and simplified examples researchers use to demonstrate their work understandably only seldom convince the industry to risk investing in AI tools. In this talk, I will speak about my experience establishing trust in AI-based tools to support creative processes in game development. Having worked on this topic in both industry and academia, I will address issues ranging from establishing a common language and explaining AI behaviour to issuing performance guarantees via benchmarking and theoretical analysis. Mike Preuss (Leiden University, The Netherlands): In the eye of the storm? Where are we going with game AI? Original talk abstract: Looking back at the last 10 years of research in Game AI we find that Big Tech research has shaken up things quite a lot. A number of challenges were resolved in record time (Go, StarCraft, etc) and AI algorithm development is probably still increasing in speed. However, it seems that the use of AI in game-making has not changed that much, and academic research often opts for "smaller problems", slowly turning towards Human-Centered AI as possibly most important general research direction. How can we approach the next leap predicted by Alex Champandard 10 years ago of really intelligent game AI? And where would we want that? Mike presents some inconclusive thoughts and ideas on future developments. The Game AI Meetup takes place several times a year. To sign up and receive updates, please register/join here: https://www.meetup.com/game-ai-meetup-gaim-of-london/ Previous 1 Mar 2023 Next

< Back Efficient solutions for an intriguing failure of llms: Long context window does not mean LLMs can analyze long sequences flawlessly Link Author(s) Peyman Hosseini, Ignacio Castro, Iacopo Ghinassi, Matthew Purver Abstract More info TBA Link

< Back On State Representations and Behavioural Modelling Methods in Reinforcement Learning Link Author(s) H Siljebrat Abstract More info TBA Link

< Back Nuffield Research Placement IGGI was taking part for the first time in the Nuffield Research Placement scheme this year: IGGI PhD researchers were supervising keen A-level students from across the country so that the students could gain an insight into scientific research work. The supervision schemes lasted two weeks per student and took place over the summer. The students benefited from the exposure to complex problems, and their involvement in the respective project may very well give them a significant boost to their prospective university applications. One of the participating IGGI PhD Researchers, Michelangelo Conserva , reported: "It was a pleasure for me to take part in the project! During the first week I did a series of seminars to introduce the student to the relevant topics; whereas in the second week we did a project on generating synthetic faces using Generative Adversarial Networks. I personally believe that this is a great initiative that will increase diversity in the next generation of researchers and I am proud of my little contribution to it. I think that mixing online and in person meetings would be great but I understand that it was not possible this year." Another participant, Nuria Peña Pérez , provided the following feedback: "Participating in the Nuffield Research Placement scheme has been a very enriching experience. During this programme, I supervised an A-level student on a project related to my work. Before starting the project, we had several discussions about the objectives to be pursued, to which the student significantly contributed with their own ideas. The project lasted two weeks, during which the student worked hard on the development of a video game for rehabilitation, a process that allowed the student to improve their technical skills while I gained supervising experience. This programme has therefore greatly benefited both of us.I think that through the Nuffield programme students can get practical experience in research topics that might otherwise be not necessarily accessible outside of academic environments or programmes. This benefits both academia, through the incorporation of external feedback and the possibility of improving communication, and students who can develop their technical and research skills and see if they enjoy these topics before committing to specific programmes. Unfortunately, this year the program took place online, which was not a problem for the project I suggested, but I can imagine many disciplines would benefit from in-person collaborations as this would give students access to physical research labs and their equipment." IGGI is planning to take part again in next year's round which will most likely be conducted in person or in a blended format. Read here to learn more about Nuffield and the Research Placement scheme. Previous 9 Oct 2021 Next

< Back Automatic Goal Discovery in Subgoal Monte Carlo Tree Search Link Author(s) D Jeurissen, MHM Winands, CF Sironi, D Perez-Liebana Abstract More info TBA Link

< Back What makes icons appealing? The role of processing fluency in predicting icon appeal in different task contexts Link Author(s) S McDougall, I Reppa, J Kulik, A Taylor Abstract More info TBA Link

< Back Automated game balancing in Ms PacMan and StarCraft using evolutionary algorithms Link Author(s) M Morosan, R Poli Abstract More info TBA Link

Themes (All) iGGi is a collaboration between Uni of York + Queen Mary Uni of London: the largest training programme worldwide for doing a PhD in digital games. iGGi Themes Game AI How might we use novel AI techniques including machine learning and decision search to create more effective and engaging game agents and understand human behaviour in games? Design & Development How might we advance game-making with new design and development insights, methods, tools, and techniques? Immersive Technology How might we advance the use and understanding of VR, XR, AR, and other immersive technologies in games? Game Audio How might we combine data, AI, and music psychology to create engaging adaptive music and sound experiences for games? Esports How might we use esports data to create immersive audience experiences, help players and teams improve, and understand human performance? Applied Games How might we design and use games to support health, learning, work, and conduct scientific research? Creative Computing How can AI tools enhance the creative process of (human) game creatives? Player Research How might we use and combine diverse methods to understand how people experience games, interact with and through them, and are affected by them? Accessibility How might we design and make games playable and inclusive to as wide a range of people as possible, regardless of background or ability? Game Data How might we analyse the big data exhausts of games to support game developers and understand people inside and outside of games?

< Back The Dark Souls of Archaeology: Recording Elden Ring Link Author(s) F Smith Nicholls, M Cook Abstract More info TBA Link

< Back Searching for an (un) stable equilibrium: experiments in training generative models without data Link Author(s) T Broad, M Grierson Abstract More info TBA Link

< Back Recurrent Neural-Linear Posterior Sampling for Nonstationary Contextual Bandits Link Author(s) A Ramesh, P Rauber, M Conserva, J Schmidhuber Abstract More info TBA Link

< Back Embodied, in-medium design of VR game motion controls using interactive supervised learning Link Author(s) C Gonzalez Diaz Abstract More info TBA Link