Julian Togelius

Aaron Dharna, Charlie Summers, Rohin Dasari et al.

This paper proposes a framework called Watts for implementing, comparing, and recombining open-ended learning (OEL) algorithms. Motivated by modularity and algorithmic flexibility, Watts atomizes the components of OEL systems to promote the study of and direct comparisons between approaches. Examining implementations of three OEL algorithms, the paper introduces the modules of the framework. The hope is for Watts to enable benchmarking and to explore new types of OEL algorithms. The repo is available at \url{https://github.com/aadharna/watts}

Muhammad U. Nasir, Sam Earle, Christopher Cleghorn et al.

Large Language Models (LLMs) have emerged as powerful tools capable of accomplishing a broad spectrum of tasks. Their abilities span numerous areas, and one area where they have made a significant impact is in the domain of code generation. Here, we propose using the coding abilities of LLMs to introduce meaningful variations to code defining neural networks. Meanwhile, Quality-Diversity (QD) algorithms are known to discover diverse and robust solutions. By merging the code-generating abilities of LLMs with the diversity and robustness of QD solutions, we introduce \texttt{LLMatic}, a Neural Architecture Search (NAS) algorithm. While LLMs struggle to conduct NAS directly through prompts, \texttt{LLMatic} uses a procedural approach, leveraging QD for prompts and network architecture to create diverse and high-performing networks. We test \texttt{LLMatic} on the CIFAR-10 and NAS-bench-201 benchmarks, demonstrating that it can produce competitive networks while evaluating just $2,000$ candidates, even without prior knowledge of the benchmark domain or exposure to any previous top-performing models for the benchmark. The open-sourced code is available in \url{https://github.com/umair-nasir14/LLMatic}.

Yangkun Chen, Joseph Suarez, Junjie Zhang et al.

We present the results of the second Neural MMO challenge, hosted at IJCAI 2022, which received 1600+ submissions. This competition targets robustness and generalization in multi-agent systems: participants train teams of agents to complete a multi-task objective against opponents not seen during training. The competition combines relatively complex environment design with large numbers of agents in the environment. The top submissions demonstrate strong success on this task using mostly standard reinforcement learning (RL) methods combined with domain-specific engineering. We summarize the competition design and results and suggest that, as an academic community, competitions may be a powerful approach to solving hard problems and establishing a solid benchmark for algorithms. We will open-source our benchmark including the environment wrapper, baselines, a visualization tool, and selected policies for further research.

Zehua Jiang, Sam Earle, Michael Cerny Green et al.

Procedural Content Generation via Reinforcement Learning (PCGRL) foregoes the need for large human-authored data-sets and allows agents to train explicitly on functional constraints, using computable, user-defined measures of quality instead of target output. We explore the application of PCGRL to 3D domains, in which content-generation tasks naturally have greater complexity and potential pertinence to real-world applications. Here, we introduce several PCGRL tasks for the 3D domain, Minecraft (Mojang Studios, 2009). These tasks will challenge RL-based generators using affordances often found in 3D environments, such as jumping, multiple dimensional movement, and gravity. We train an agent to optimize each of these tasks to explore the capabilities of previous research in PCGRL. This agent is able to generate relatively complex and diverse levels, and generalize to random initial states and control targets. Controllability tests in the presented tasks demonstrate their utility to analyze success and failure for 3D generators.

Pittawat Taveekitworachai, Febri Abdullah, Mury F. Dewantoro et al.

This paper presents the first ChatGPT4PCG Competition at the 2023 IEEE Conference on Games. The objective of this competition is for participants to create effective prompts for ChatGPT--enabling it to generate Science Birds levels with high stability and character-like qualities--fully using their creativity as well as prompt engineering skills. ChatGPT is a conversational agent developed by OpenAI. Science Birds is selected as the competition platform because designing an Angry Birds-like level is not a trivial task due to the in-game gravity; the quality of the levels is determined by their stability. To lower the entry barrier to the competition, we limit the task to the generation of capitalized English alphabetical characters. We also allow only a single prompt to be used for generating all the characters. Here, the quality of the generated levels is determined by their stability and similarity to the given characters. A sample prompt is provided to participants for their reference. An experiment is conducted to determine the effectiveness of several modified versions of this sample prompt on level stability and similarity by testing them on several characters. To the best of our knowledge, we believe that ChatGPT4PCG is the first competition of its kind and hope to inspire enthusiasm for prompt engineering in procedural content generation.

Debosmita Bhaumik, Ahmed Khalifa, Julian Togelius

We present Lode Encoder, a gamified mixed-initiative level creation system for the classic platform-puzzle game Lode Runner. The system is built around several autoencoders which are trained on sets of Lode Runner levels. When fed with the user's design, each autoencoder produces a version of that design which is closer in style to the levels that it was trained on. The Lode Encoder interface allows the user to build and edit levels through 'painting' from the suggestions provided by the autoencoders. Crucially, in order to encourage designers to explore new possibilities, the system does not include more traditional editing tools. We report on the system design and training procedure, as well as on the evolution of the system itself and user tests.

Alberto Alvarez, Jose Font, Julian Togelius

Narratives are a predominant part of games, and their design poses challenges when identifying, encoding, interpreting, evaluating, and generating them. One way to address this would be to approach narrative design in a more abstract layer, such as narrative structures. This paper presents Story Designer, a mixed-initiative co-creative narrative structure tool built on top of the Evolutionary Dungeon Designer (EDD) that uses tropes, narrative conventions found across many media types, to design these structures. Story Designer uses tropes as building blocks for narrative designers to compose complete narrative structures by interconnecting them in graph structures called narrative graphs. Our mixed-initiative approach lets designers manually create their narrative graphs and feeds an underlying evolutionary algorithm with those, creating quality-diverse suggestions using MAP-Elites. Suggestions are visually represented for designers to compare and evaluate and can then be incorporated into the design for further manual editions. At the same time, we use the levels designed within EDD as constraints for the narrative structure, intertwining both level design and narrative. We evaluate the impact of these constraints and the system's adaptability and expressiveness, resulting in a potential tool to create narrative structures combining level design aspects with narrative.

Ahmed Khalifa, Michael Cerny Green, Julian Togelius

Search-based procedural content generation (PCG) is a well-known method for level generation in games. Its key advantage is that it is generic and able to satisfy functional constraints. However, due to the heavy computational costs to run these algorithms online, search-based PCG is rarely utilized for real-time generation. In this paper, we introduce mutation models, a new type of iterative level generator based on machine learning. We train a model to imitate the evolutionary process and use the trained model to generate levels. This trained model is able to modify noisy levels sequentially to create better levels without the need for a fitness function during inference. We evaluate our trained models on a 2D maze generation task. We compare several different versions of the method: training the models either at the end of evolution (normal evolution) or every 100 generations (assisted evolution) and using the model as a mutation function during evolution. Using the assisted evolution process, the final trained models are able to generate mazes with a success rate of 99% and high diversity of 86%. The trained model is many times faster than the evolutionary process it was trained on. This work opens the door to a new way of learning level generators guided by an evolutionary process, meaning automatic creation of generators with specifiable constraints and objectives that are fast enough for runtime deployment in games.

M Charity, Julian Togelius

The Keke AI Competition introduces an artificial agent competition for the game Baba is You - a Sokoban-like puzzle game where players can create rules that influence the mechanics of the game. Altering a rule can cause temporary or permanent effects for the rest of the level that could be part of the solution space. The nature of these dynamic rules and the deterministic aspect of the game creates a challenge for AI to adapt to a variety of mechanic combinations in order to solve a level. This paper describes the framework and evaluation metrics used to rank submitted agents and baseline results from sample tree search agents.

M Charity, Yash Bhartia, Daniel Zhang et al.

This paper introduces a system used to generate game feature suggestions based on a text prompt. Trained on the game descriptions of almost 60k games, it uses the word embeddings of a small GLoVe model to extract features and entities found in thematically similar games which are then passed through a generator model to generate new features for a user's prompt. We perform a short user study comparing the features generated from a fine-tuned GPT-2 model, a model using the ConceptNet, and human-authored game features. Although human suggestions won the overall majority of votes, the GPT-2 model outperformed the human suggestions in certain games. This system is part of a larger game design assistant tool that is able to collaborate with users at a conceptual level.

Aaron Dharna, Amy K Hoover, Julian Togelius et al.

PINSKY is a system for open-ended learning through neuroevolution in game-based domains. It builds on the Paired Open-Ended Trailblazer (POET) system, which originally explored learning and environment generation for bipedal walkers, and adapts it to games in the General Video Game AI (GVGAI) system. Previous work showed that by co-evolving levels and neural network policies, levels could be found for which successful policies could not be created via optimization alone. Studied in the realm of Artificial Life as a potentially open-ended alternative to gradient-based fitness, minimal criteria (MC)-based selection helps foster diversity in evolutionary populations. The main question addressed by this paper is how the open-ended learning actually works, focusing in particular on the role of transfer of policies from one evolutionary branch ("species") to another. We analyze the dynamics of the system through creating phylogenetic trees, analyzing evolutionary trajectories of policies, and temporally breaking down transfers according to species type. Furthermore, we analyze the impact of the minimal criterion on generated level diversity and inter-species transfer. The most insightful finding is that inter-species transfer, while rare, is crucial to the system's success.

Anubhav Jain, Nasir Memon, Julian Togelius

Face swapping technology used to create "Deepfakes" has advanced significantly over the past few years and now enables us to create realistic facial manipulations. Current deep learning algorithms to detect deepfakes have shown promising results, however, they require large amounts of training data, and as we show they are biased towards a particular ethnicity. We propose a deepfake detection methodology that eliminates the need for any real data by making use of synthetically generated data using StyleGAN3. This not only performs at par with the traditional training methodology of using real data but it shows better generalization capabilities when finetuned with a small amount of real data. Furthermore, this also reduces biases created by facial image datasets that might have sparse data from particular ethnicities.

M Charity, Nasir Memon, Zehua Jiang et al.

This work expands on previous advancements in genetic fingerprint spoofing via the DeepMasterPrints and introduces Diversity and Novelty MasterPrints. This system uses quality diversity evolutionary algorithms to generate dictionaries of artificial prints with a focus on increasing coverage of users from the dataset. The Diversity MasterPrints focus on generating solution prints that match with users not covered by previously found prints, and the Novelty MasterPrints explicitly search for prints with more that are farther in user space than previous prints. Our multi-print search methodologies outperform the singular DeepMasterPrints in both coverage and generalization while maintaining quality of the fingerprint image output.

M Charity, Julian Togelius

This paper describes the implementation of the Aesthetic Bot, an automated Twitter account that posts images of small game maps that are either user-made or generated from an evolutionary system. The bot then prompts users to vote via a poll posted in the image's thread for the most aesthetically pleasing map. This creates a rating system that allows for direct interaction with the bot in a way that is integrated seamlessly into a user's regularly updated Twitter content feed. Upon conclusion of the each voting round, the bot learns from the distribution of votes for each map to emulate user preferences for design and visual aesthetic in order to generate maps that would win future vote pairings. We discuss the ongoing results and emerging behaviors that have occurred since the release of this system from both the bot's generation of game maps and the participating Twitter users.

Michael Cerny Green, Ahmed Khalifa, M Charity et al.

We investigate how to efficiently predict play personas based on playtraces. Play personas can be computed by calculating the action agreement ratio between a player and a generative model of playing behavior, a so-called procedural persona. But this is computationally expensive and assumes that appropriate procedural personas are readily available. We present two methods for estimating player persona, one using regular supervised learning and aggregate measures of game mechanics initiated, and another based on sequence learning on a trace of closely cropped gameplay observations. While both of these methods achieve high accuracy when predicting play personas defined by agreement with procedural personas, they utterly fail to predict play style as defined by the players themselves using a questionnaire. This interesting result highlights the value of using computational methods in defining play personas.

Shuo Huang, Chengpeng Hu, Julian Togelius et al.

Procedurally generating cities in Minecraft provides players more diverse scenarios and could help understand and improve the design of cities in other digital worlds and the real world. This paper presents a city generator that was submitted as an entry to the 2023 Edition of Minecraft Settlement Generation Competition for Minecraft. The generation procedure is composed of six main steps, namely vegetation clearing, terrain reshaping, building layout generation, route planning, streetlight placement, and wall construction. Three algorithms, including a heuristic-based algorithm, an evolving layout algorithm, and a random one are applied to generate the building layout, thus determining where to place different redstone style buildings, and tested by generating cities on random maps in limited time. Experimental results show that the heuristic-based algorithm is capable of finding an acceptable building layout faster for flat maps, while the evolving layout algorithm performs better in evolving layout for rugged maps. A user study is conducted to compare our generator with outstanding entries of the competition's 2022 edition using the competition's evaluation criteria and shows that our generator performs well in the adaptation and functionality criteria

Graham Todd, Sam Earle, Muhammad Umair Nasir et al.

Large Language Models (LLMs) are powerful tools, capable of leveraging their training on natural language to write stories, generate code, and answer questions. But can they generate functional video game levels? Game levels, with their complex functional constraints and spatial relationships in more than one dimension, are very different from the kinds of data an LLM typically sees during training. Datasets of game levels are also hard to come by, potentially taxing the abilities of these data-hungry models. We investigate the use of LLMs to generate levels for the game Sokoban, finding that LLMs are indeed capable of doing so, and that their performance scales dramatically with dataset size. We also perform preliminary experiments on controlling LLM level generators and discuss promising areas for future work.

Debosmita Bhaumik, Julian Togelius, Georgios N. Yannakakis et al.

We explore AI-powered upscaling as a design assistance tool in the context of creating 2D game levels. Deep neural networks are used to upscale artificially downscaled patches of levels from the puzzle platformer game Lode Runner. The trained networks are incorporated into a web-based editor, where the user can create and edit levels at three different levels of resolution: 4x4, 8x8, and 16x16. An edit at any resolution instantly transfers to the other resolutions. As upscaling requires inventing features that might not be present at lower resolutions, we train neural networks to reproduce these features. We introduce a neural network architecture that is capable of not only learning upscaling but also giving higher priority to less frequent tiles. To investigate the potential of this tool and guide further development, we conduct a qualitative study with 3 designers to understand how they use it. Designers enjoyed co-designing with the tool, liked its underlying concept, and provided feedback for further improvement.

Michael Cerny Green, Ahmed Khalifa, M Charity et al.

In this paper, we present a method for automated persona-driven video game tutorial level generation. Tutorial levels are scenarios in which the player can explore and discover different rules and game mechanics. Procedural personas can guide generators to create content which encourages or discourages certain playstyle behaviors. In this system, we use procedural personas to calculate the behavioral characteristics of levels which are evolved using the quality-diversity algorithm known as Constrained MAP-Elites. An evolved map's quality is determined by its simplicity: the simpler it is, the better it is. Within this work, we show that the generated maps can strongly encourage or discourage different persona-like behaviors and range from simple solutions to complex puzzle-levels, making them perfect candidates for a tutorial generative system.

Ya-Chuan Hsu, Matthew C. Fontaine, Sam Earle et al.

We present a method for generating arrangements of indoor furniture from human-designed furniture layout data. Our method creates arrangements that target specified diversity, such as the total price of all furniture in the room and the number of pieces placed. To generate realistic furniture arrangement, we train a generative adversarial network (GAN) on human-designed layouts. To target specific diversity in the arrangements, we optimize the latent space of the GAN via a quality diversity algorithm to generate a diverse arrangement collection. Experiments show our approach discovers a set of arrangements that are similar to human-designed layouts but varies in price and number of furniture pieces.

Anubhav Jain, Nasir Memon, Julian Togelius

Deep generative models require large amounts of training data. This often poses a problem as the collection of datasets can be expensive and difficult, in particular datasets that are representative of the appropriate underlying distribution (e.g. demographic). This introduces biases in datasets which are further propagated in the models. We present an approach to construct an unbiased generative adversarial network (GAN) from an existing biased GAN by rebalancing the model distribution. We do so by generating balanced data from an existing imbalanced deep generative model using an evolutionary algorithm and then using this data to train a balanced generative model. Additionally, we propose a bias mitigation loss function that minimizes the deviation of the learned class distribution from being equiprobable. We show results for the StyleGAN2 models while training on the Flickr Faces High Quality (FFHQ) dataset for racial fairness and see that the proposed approach improves on the fairness metric by almost 5 times, whilst maintaining image quality. We further validate our approach by applying it to an imbalanced CIFAR10 dataset where we show that we can obtain comparable fairness and image quality as when training on a balanced CIFAR10 dataset which is also twice as large. Lastly, we argue that the traditionally used image quality metrics such as Frechet inception distance (FID) are unsuitable for scenarios where the class distributions are imbalanced and a balanced reference set is not available.

Tim Merino, Sam Earle, Ryunosuke Iwai et al.

We introduce Dream-Cubed, a large-scale dataset of Minecraft worlds at voxel resolution, and a family of models using cubes as powerful compositional units for efficient generation of interactive 3D environments. Dream-Cubed comprises tens of billions of tokens from a carefully curated mixture of procedural biome terrain and high-quality human-authored maps. We use this dataset to conduct the first large-scale study of 3D diffusion models for voxel generation, analyzing discrete and continuous diffusion formulations, data compositions, and architectural design choices. Our models operate directly in the space of blocks, enabling efficient and semantically grounded generation while supporting interactive user workflows such as inpainting and outpainting from user-authored blocks. To quantitatively evaluate our models, we adapt the FID metric to assess semantic differences between real and generated world renderings, and validate generation quality through a human preference study. We release the full dataset, code, and all our pretrained models, which we hope will provide a foundation for future research in efficient generative modeling for structured, interactive 3D environments.

Graham Todd, Alexander Padula, Matthew Stephenson et al.

Automatically generating novel and interesting games is a complex task. Challenges include representing game rules in a computationally workable form, searching through the large space of potential games under most such representations, and accurately evaluating the originality and quality of previously unseen games. Prior work in automated game generation has largely focused on relatively restricted rule representations and relied on domain-specific heuristics. In this work, we explore the generation of novel games in the comparatively expansive Ludii game description language, which encodes the rules of over 1000 board games in a variety of styles and modes of play. We draw inspiration from recent advances in large language models and evolutionary computation in order to train a model that intelligently mutates and recombines games and mechanics expressed as code. We demonstrate both quantitatively and qualitatively that our approach is capable of generating new and interesting games, including in regions of the potential rules space not covered by existing games in the Ludii dataset. A sample of the generated games are available to play online through the Ludii portal.

Enhong Liu, Joseph Suarez, Chenhui You et al.

In this paper, we present the results of the NeurIPS-2022 Neural MMO Challenge, which attracted 500 participants and received over 1,600 submissions. Like the previous IJCAI-2022 Neural MMO Challenge, it involved agents from 16 populations surviving in procedurally generated worlds by collecting resources and defeating opponents. This year's competition runs on the latest v1.6 Neural MMO, which introduces new equipment, combat, trading, and a better scoring system. These elements combine to pose additional robustness and generalization challenges not present in previous competitions. This paper summarizes the design and results of the challenge, explores the potential of this environment as a benchmark for learning methods, and presents some practical reinforcement learning training approaches for complex tasks with sparse rewards. Additionally, we have open-sourced our baselines, including environment wrappers, benchmarks, and visualization tools for future research.

Sam Earle, Ozlem Yildiz, Julian Togelius et al.

Pathfinding makes up an important sub-component of a broad range of complex tasks in AI, such as robot path planning, transport routing, and game playing. While classical algorithms can efficiently compute shortest paths, neural networks could be better suited to adapting these sub-routines to more complex and intractable tasks. As a step toward developing such networks, we hand-code and learn models for Breadth-First Search (BFS), i.e. shortest path finding, using the unified architectural framework of Neural Cellular Automata, which are iterative neural networks with equal-size inputs and outputs. Similarly, we present a neural implementation of Depth-First Search (DFS), and outline how it can be combined with neural BFS to produce an NCA for computing diameter of a graph. We experiment with architectural modifications inspired by these hand-coded NCAs, training networks from scratch to solve the diameter problem on grid mazes while exhibiting strong generalization ability. Finally, we introduce a scheme in which data points are mutated adversarially during training. We find that adversarially evolving mazes leads to increased generalization on out-of-distribution examples, while at the same time generating data-sets with significantly more complex solutions for reasoning tasks.

Timothy Merino, Roman Negri, Dipika Rajesh et al.

The five-dollar model is a lightweight text-to-image generative architecture that generates low dimensional images from an encoded text prompt. This model can successfully generate accurate and aesthetically pleasing content in low dimensional domains, with limited amounts of training data. Despite the small size of both the model and datasets, the generated images are still able to maintain the encoded semantic meaning of the textual prompt. We apply this model to three small datasets: pixel art video game maps, video game sprite images, and down-scaled emoji images and apply novel augmentation strategies to improve the performance of our model on these limited datasets. We evaluate our models performance using cosine similarity score between text-image pairs generated by the CLIP VIT-B/32 model.

Julian Togelius, Ahmed Khalifa, Sam Earle et al.

Evolutionary machine learning (EML) has been applied to games in multiple ways, and for multiple different purposes. Importantly, AI research in games is not only about playing games; it is also about generating game content, modeling players, and many other applications. Many of these applications pose interesting problems for EML. We will structure this chapter on EML for games based on whether evolution is used to augment machine learning (ML) or ML is used to augment evolution. For completeness, we also briefly discuss the usage of ML and evolution separately in games.

Sam Earle, Zehua Jiang, Julian Togelius

Procedural Content Generation via Reinforcement Learning (PCGRL) has been introduced as a means by which controllable designer agents can be trained based only on a set of computable metrics acting as a proxy for the level's quality and key characteristics. While PCGRL offers a unique set of affordances for game designers, it is constrained by the compute-intensive process of training RL agents, and has so far been limited to generating relatively small levels. To address this issue of scale, we implement several PCGRL environments in Jax so that all aspects of learning and simulation happen in parallel on the GPU, resulting in faster environment simulation; removing the CPU-GPU transfer of information bottleneck during RL training; and ultimately resulting in significantly improved training speed. We replicate several key results from prior works in this new framework, letting models train for much longer than previously studied, and evaluating their behavior after 1 billion timesteps. Aiming for greater control for human designers, we introduce randomized level sizes and frozen "pinpoints" of pivotal game tiles as further ways of countering overfitting. To test the generalization ability of learned generators, we evaluate models on large, out-of-distribution map sizes, and find that partial observation sizes learn more robust design strategies.

Tim Merino, Sam Earle, Ryan Sudhakaran et al.

The Connections puzzle is a word association game published daily by The New York Times (NYT). In this game, players are asked to find groups of four words that are connected by a common theme. While solving a given Connections puzzle requires both semantic knowledge and abstract reasoning, generating novel puzzles additionally requires a form of metacognition: generators must be able to accurately model the downstream reasoning of potential solvers. In this paper, we investigate the ability of the GPT family of Large Language Models (LLMs) to generate challenging and creative word games for human players. We start with an analysis of the word game Connections and the unique challenges it poses as a Procedural Content Generation (PCG) domain. We then propose a method for generating Connections puzzles using LLMs by adapting a Tree of Thoughts (ToT) prompting approach. We evaluate this method by conducting a user study, asking human players to compare AI-generated puzzles against published Connections puzzles. Our findings show that LLMs are capable puzzle creators, and can generate diverse sets of enjoyable, challenging, and creative Connections puzzles as judged by human users.

M Charity, Dipika Rajesh, Sam Earle et al.

We introduce a system called Amorphous Fortress -- an abstract, yet spatial, open-ended artificial life simulation. In this environment, the agents are represented as finite-state machines (FSMs) which allow for multi-agent interaction within a constrained space. These agents are created by randomly generating and evolving the FSMs; sampling from pre-defined states and transitions. This environment was designed to explore the emergent AI behaviors found implicitly in simulation games such as Dwarf Fortress or The Sims. We apply the hill-climber evolutionary search algorithm to this environment to explore the various levels of depth and interaction from the generated FSMs.

Sam Earle, Julian Togelius

We introduce Autoverse, an evolvable, domain-specific language for single-player 2D grid-based games, and demonstrate its use as a scalable training ground for Open-Ended Learning (OEL) algorithms. Autoverse uses cellular-automaton-like rewrite rules to describe game mechanics, allowing it to express various game environments (e.g. mazes, dungeons, sokoban puzzles) that are popular testbeds for Reinforcement Learning (RL) agents. Each rewrite rule can be expressed as a series of simple convolutions, allowing for environments to be parallelized on the GPU, thereby drastically accelerating RL training. Using Autoverse, we propose jump-starting open-ended learning by imitation learning from search. In such an approach, we first evolve Autoverse environments (their rules and initial map topology) to maximize the number of iterations required by greedy tree search to discover a new best solution, producing a curriculum of increasingly complex environments and playtraces. We then distill these expert playtraces into a neural-network-based policy using imitation learning. Finally, we use the learned policy as a starting point for open-ended RL, where new training environments are continually evolved to maximize the RL player agent's value function error (a proxy for its regret, or the learnability of generated environments), finding that this approach improves the performance and generality of resultant player agents.

Sudipta Banerjee, Anubhav Jain, Zehua Jiang et al.

A dictionary attack in a biometric system entails the use of a small number of strategically generated images or templates to successfully match with a large number of identities, thereby compromising security. We focus on dictionary attacks at the template level, specifically the IrisCodes used in iris recognition systems. We present an hitherto unknown vulnerability wherein we mix IrisCodes using simple bitwise operators to generate alpha-mixtures - alpha-wolves (combining a set of "wolf" samples) and alpha-mammals (combining a set of users selected via search optimization) that increase false matches. We evaluate this vulnerability using the IITD, CASIA-IrisV4-Thousand and Synthetic datasets, and observe that an alpha-wolf (from two wolves) can match upto 71 identities @FMR=0.001%, while an alpha-mammal (from two identities) can match upto 133 other identities @FMR=0.01% on the IITD dataset.

Graham Todd, Alexander G. Padula, Dennis J. N. J. Soemers et al.

Games have long been used as benchmarks and testing environments for research in artificial intelligence. A key step in supporting this research was the development of game description languages: frameworks that compile domain-specific code into playable and simulatable game environments, allowing researchers to generalize their algorithms and approaches across multiple games without having to manually implement each one. More recently, progress in reinforcement learning (RL) has been largely driven by advances in hardware acceleration. Libraries like JAX allow practitioners to take full advantage of cutting-edge computing hardware, often speeding up training and testing by orders of magnitude. Here, we present a synthesis of these strands of research: a domain-specific language for board games which automatically compiles into hardware-accelerated code. Our framework, Ludax, combines the generality of game description languages with the speed of modern parallel processing hardware and is designed to fit neatly into existing deep learning pipelines. We envision Ludax as a tool to help accelerate games research generally, from RL to cognitive science, by enabling rapid simulation and providing a flexible representation scheme. We present a detailed breakdown of Ludax's description language and technical notes on the compilation process, along with speed benchmarking and a demonstration of training RL agents. The Ludax framework, along with implementations of existing board games, is open-source and freely available.

Yuhe Nie, Michael Middleton, Tim Merino et al.

Procedural Content Generation via Machine Learning (PCGML) has enhanced game content creation, yet challenges in controllability and limited training data persist. This study addresses these issues by distilling a constructive PCG algorithm into a controllable PCGML model. We first generate a large amount of content with a constructive algorithm and label it using a Large Language Model (LLM). We use these synthetic labels to condition two PCGML models for content-specific generation, a diffusion model and the five-dollar model. This neural network distillation process ensures that the generation aligns with the original algorithm while introducing controllability through plain text. We define this text-conditioned PCGML as a Text-to-game-Map (T2M) task, offering an alternative to prevalent text-to-image multi-modal tasks. We compare our distilled models with the baseline constructive algorithm. Our analysis of the variety, accuracy, and quality of our generation demonstrates the efficacy of distilling constructive methods into controllable text-conditioned PCGML models.

Shuo Huang, Muhammad Umair Nasir, Steven James et al.

We present Word2Minecraft, a system that leverages large language models to generate playable game levels in Minecraft based on structured stories. The system transforms narrative elements-such as protagonist goals, antagonist challenges, and environmental settings-into game levels with both spatial and gameplay constraints. We introduce a flexible framework that allows for the customization of story complexity, enabling dynamic level generation. The system employs a scaling algorithm to maintain spatial consistency while adapting key game elements. We evaluate Word2Minecraft using both metric-based and human-based methods. Our results show that GPT-4-Turbo outperforms GPT-4o-Mini in most areas, including story coherence and objective enjoyment, while the latter excels in aesthetic appeal. We also demonstrate the system' s ability to generate levels with high map enjoyment, offering a promising step forward in the intersection of story generation and game design. We open-source the code at https://github.com/JMZ-kk/Word2Minecraft/tree/word2mc_v0

Roberto Gallotta, Graham Todd, Marvin Zammit et al.

Recent years have seen an explosive increase in research on large language models (LLMs), and accompanying public engagement on the topic. While starting as a niche area within natural language processing, LLMs have shown remarkable potential across a broad range of applications and domains, including games. This paper surveys the current state of the art across the various applications of LLMs in and for games, and identifies the different roles LLMs can take within a game. Importantly, we discuss underexplored areas and promising directions for future uses of LLMs in games and we reconcile the potential and limitations of LLMs within the games domain. As the first comprehensive survey and roadmap at the intersection of LLMs and games, we are hopeful that this paper will serve as the basis for groundbreaking research and innovation in this exciting new field.

Muhammad U. Nasir, Steven James, Julian Togelius

Large Language Models (LLMs) have proven their worth across a diverse spectrum of disciplines. LLMs have shown great potential in Procedural Content Generation (PCG) as well, but directly generating a level through a pre-trained LLM is still challenging. This work introduces Word2World, a system that enables LLMs to procedurally design playable games through stories, without any task-specific fine-tuning. Word2World leverages the abilities of LLMs to create diverse content and extract information. Combining these abilities, LLMs can create a story for the game, design narrative, and place tiles in appropriate places to create coherent worlds and playable games. We test Word2World with different LLMs and perform a thorough ablation study to validate each step. We open-source the code at https://github.com/umair-nasir14/Word2World.

Bryon Tjanaka, Matthew C. Fontaine, Julian Togelius et al.

Consider the problem of training robustly capable agents. One approach is to generate a diverse collection of agent polices. Training can then be viewed as a quality diversity (QD) optimization problem, where we search for a collection of performant policies that are diverse with respect to quantified behavior. Recent work shows that differentiable quality diversity (DQD) algorithms greatly accelerate QD optimization when exact gradients are available. However, agent policies typically assume that the environment is not differentiable. To apply DQD algorithms to training agent policies, we must approximate gradients for performance and behavior. We propose two variants of the current state-of-the-art DQD algorithm that compute gradients via approximation methods common in reinforcement learning (RL). We evaluate our approach on four simulated locomotion tasks. One variant achieves results comparable to the current state-of-the-art in combining QD and RL, while the other performs comparably in two locomotion tasks. These results provide insight into the limitations of current DQD algorithms in domains where gradients must be approximated. Source code is available at https://github.com/icaros-usc/dqd-rl

Diego Perez-Liebana, Jialin Liu, Ahmed Khalifa et al.

General Video Game Playing (GVGP) aims at designing an agent that is capable of playing multiple video games with no human intervention. In 2014, The General Video Game AI (GVGAI) competition framework was created and released with the purpose of providing researchers a common open-source and easy to use platform for testing their AI methods with potentially infinity of games created using Video Game Description Language (VGDL). The framework has been expanded into several tracks during the last few years to meet the demand of different research directions. The agents are required either to play multiple unknown games with or without access to game simulations, or to design new game levels or rules. This survey paper presents the VGDL, the GVGAI framework, existing tracks, and reviews the wide use of GVGAI framework in research, education and competitions five years after its birth. A future plan of framework improvements is also described.

Asad Anjum, Yuting Li, Noelle Law et al.

This paper studies how large language models (LLMs) can act as effective, high-level creative collaborators and ``muses'' for game design. We model the design of this study after the exercises artists use by looking at amorphous ink splotches for creative inspiration. Our goal is to determine whether AI-assistance can improve, hinder, or provide an alternative quality to games when compared to the creative intents implemented by human designers. The capabilities of LLMs as game designers are stress tested by placing it at the forefront of the decision making process. Three prototype games are designed across 3 different genres: (1) a minimalist base game, (2) a game with features and game feel elements added by a human game designer, and (3) a game with features and feel elements directly implemented from prompted outputs of the LLM, ChatGPT. A user study was conducted and participants were asked to blindly evaluate the quality and their preference of these games. We discuss both the development process of communicating creative intent to an AI chatbot and the synthesized open feedback of the participants. We use this data to determine both the benefits and shortcomings of AI in a more design-centric role.

Sam Earle, Filippos Kokkinos, Yuhe Nie et al.

Procedural Content Generation (PCG) algorithms enable the automatic generation of complex and diverse artifacts. However, they don't provide high-level control over the generated content and typically require domain expertise. In contrast, text-to-3D methods allow users to specify desired characteristics in natural language, offering a high amount of flexibility and expressivity. But unlike PCG, such approaches cannot guarantee functionality, which is crucial for certain applications like game design. In this paper, we present a method for generating functional 3D artifacts from free-form text prompts in the open-world game Minecraft. Our method, DreamCraft, trains quantized Neural Radiance Fields (NeRFs) to represent artifacts that, when viewed in-game, match given text descriptions. We find that DreamCraft produces more aligned in-game artifacts than a baseline that post-processes the output of an unconstrained NeRF. Thanks to the quantized representation of the environment, functional constraints can be integrated using specialized loss terms. We show how this can be leveraged to generate 3D structures that match a target distribution or obey certain adjacency rules over the block types. DreamCraft inherits a high degree of expressivity and controllability from the NeRF, while still being able to incorporate functional constraints through domain-specific objectives.

Anubhav Jain, Yuya Kobayashi, Takashi Shibuya et al.

Recent advancements in text-to-image diffusion models have brought them to the public spotlight, becoming widely accessible and embraced by everyday users. However, these models have been shown to generate harmful content such as not-safe-for-work (NSFW) images. While approaches have been proposed to erase such abstract concepts from the models, jail-breaking techniques have succeeded in bypassing such safety measures. In this paper, we propose TraSCE, an approach to guide the diffusion trajectory away from generating harmful content. Our approach is based on negative prompting, but as we show in this paper, a widely used negative prompting strategy is not a complete solution and can easily be bypassed in some corner cases. To address this issue, we first propose using a specific formulation of negative prompting instead of the widely used one. Furthermore, we introduce a localized loss-based guidance that enhances the modified negative prompting technique by steering the diffusion trajectory. We demonstrate that our proposed method achieves state-of-the-art results on various benchmarks in removing harmful content, including ones proposed by red teams, and erasing artistic styles and objects. Our proposed approach does not require any training, weight modifications, or training data (either image or prompt), making it easier for model owners to erase new concepts.

Graham Todd, Tim Merino, Sam Earle et al.

The Connections puzzle published each day by the New York Times tasks players with dividing a bank of sixteen words into four groups of four words that each relate to a common theme. Solving the puzzle requires both common linguistic knowledge (i.e. definitions and typical usage) as well as, in many cases, lateral or abstract thinking. This is because the four categories ascend in complexity, with the most challenging category often requiring thinking about words in uncommon ways or as parts of larger phrases. We investigate the capacity for automated AI systems to play Connections and explore the game's potential as an automated benchmark for abstract reasoning and a way to measure the semantic information encoded by data-driven linguistic systems. In particular, we study both a sentence-embedding baseline and modern large language models (LLMs). We report their accuracy on the task, measure the impacts of chain-of-thought prompting, and discuss their failure modes. Overall, we find that the Connections task is challenging yet feasible, and a strong test-bed for future work.

Anubhav Jain, Yuya Kobayashi, Naoki Murata et al.

Watermarking techniques are vital for protecting intellectual property and preventing fraudulent use of media. Most previous watermarking schemes designed for diffusion models embed a secret key in the initial noise. The resulting pattern is often considered hard to remove and forge into unrelated images. In this paper, we propose a black-box adversarial attack without presuming access to the diffusion model weights. Our attack uses only a single watermarked example and is based on a simple observation: there is a many-to-one mapping between images and initial noises. There are regions in the clean image latent space pertaining to each watermark that get mapped to the same initial noise when inverted. Based on this intuition, we propose an adversarial attack to forge the watermark by introducing perturbations to the images such that we can enter the region of watermarked images. We show that we can also apply a similar approach for watermark removal by learning perturbations to exit this region. We report results on multiple watermarking schemes (Tree-Ring, RingID, WIND, and Gaussian Shading) across two diffusion models (SDv1.4 and SDv2.0). Our results demonstrate the effectiveness of the attack and expose vulnerabilities in the watermarking methods, motivating future research on improving them.

Anubhav Jain, Yuya Kobayashi, Takashi Shibuya et al.

Diffusion models are prone to exactly reproduce images from the training data. This exact reproduction of the training data is concerning as it can lead to copyright infringement and/or leakage of privacy-sensitive information. In this paper, we present a novel perspective on the memorization phenomenon and propose a simple yet effective approach to mitigate it. We argue that memorization occurs because of an attraction basin in the denoising process which steers the diffusion trajectory towards a memorized image. However, this can be mitigated by guiding the diffusion trajectory away from the attraction basin by not applying classifier-free guidance until an ideal transition point occurs from which classifier-free guidance is applied. This leads to the generation of non-memorized images that are high in image quality and well-aligned with the conditioning mechanism. To further improve on this, we present a new guidance technique, opposite guidance, that escapes the attraction basin sooner in the denoising process. We demonstrate the existence of attraction basins in various scenarios in which memorization occurs, and we show that our proposed approach successfully mitigates memorization.

In-Chang Baek, Sung-Hyun Kim, Sam Earle et al.

Reward design plays a pivotal role in the training of game AIs, requiring substantial domain-specific knowledge and human effort. In recent years, several studies have explored reward generation for training game agents and controlling robots using large language models (LLMs). In the content generation literature, there has been early work on generating reward functions for reinforcement learning agent generators. This work introduces PCGRLLM, an extended architecture based on earlier work, which employs a feedback mechanism and several reasoning-based prompt engineering techniques. We evaluate the proposed method on a story-to-reward generation task in a two-dimensional environment using two state-of-the-art LLMs, demonstrating the generalizability of our approach. Our experiments provide insightful evaluations that demonstrate the capabilities of LLMs essential for content generation tasks. The results highlight significant performance improvements of 415% and 40% respectively, depending on the zero-shot capabilities of the language model. Our work demonstrates the potential to reduce human dependency in game AI development, while supporting and enhancing creative processes.

Pittawat Taveekitworachai, Febri Abdullah, Mury F. Dewantoro et al.

This paper presents the second ChatGPT4PCG competition at the 2024 IEEE Conference on Games. In this edition of the competition, we follow the first edition, but make several improvements and changes. We introduce a new evaluation metric along with allowing a more flexible format for participants' submissions and making several improvements to the evaluation pipeline. Continuing from the first edition, we aim to foster and explore the realm of prompt engineering (PE) for procedural content generation (PCG). While the first competition saw success, it was hindered by various limitations; we aim to mitigate these limitations in this edition. We introduce diversity as a new metric to discourage submissions aimed at producing repetitive structures. Furthermore, we allow submission of a Python program instead of a prompt text file for greater flexibility in implementing advanced PE approaches, which may require control flow, including conditions and iterations. We also make several improvements to the evaluation pipeline with a better classifier for similarity evaluation and better-performing function signatures. We thoroughly evaluate the effectiveness of the new metric and the improved classifier. Additionally, we perform an ablation study to select a function signature to instruct ChatGPT for level generation. Finally, we provide implementation examples of various PE techniques in Python and evaluate their preliminary performance. We hope this competition serves as a resource and platform for learning about PE and PCG in general.

Yuchen Li, Cong Lin, Muhammad Umair Nasir et al.

We introduce GVGAI-LLM, a video game benchmark for evaluating the reasoning and problem-solving capabilities of large language models (LLMs). Built on the General Video Game AI framework, it features a diverse collection of arcade-style games designed to test a model's ability to handle tasks that differ from most existing LLM benchmarks. The benchmark leverages a game description language that enables rapid creation of new games and levels, helping to prevent overfitting over time. Each game scene is represented by a compact set of ASCII characters, allowing for efficient processing by language models. GVGAI-LLM defines interpretable metrics, including the meaningful step ratio, step efficiency, and overall score, to assess model behavior. Through zero-shot evaluations across a broad set of games and levels with diverse challenges and skill depth, we reveal persistent limitations of LLMs in spatial reasoning and basic planning. Current models consistently exhibit spatial and logical errors, motivating structured prompting and spatial grounding techniques. While these interventions lead to partial improvements, the benchmark remains very far from solved. GVGAI-LLM provides a reproducible testbed for advancing research on language model capabilities, with a particular emphasis on agentic behavior and contextual reasoning.

Yunge Wen, Chenliang Huang, Hangyu Zhou et al.

The emotional arc is a universal narrative structure underlying stories across cultures and media -- an idea central to structuralist narratology, often encapsulated in the phrase "all stories are one story." We present a framework for procedural game narrative generation that incorporates emotional arcs as a structural backbone for both story progression and gameplay dynamics. Leveraging established narratological theories and large-scale empirical analyses, we focus on two core emotional patterns -- Rise and Fall -- to guide the generation of branching story graphs. Each story node is automatically populated with characters, items, and gameplay-relevant attributes (e.g., health, attack), with difficulty adjusted according to the emotional trajectory. Implemented in a prototype action role-playing game (ARPG), our system demonstrates how emotional arcs can be operationalized using large language models (LLMs) and adaptive entity generation. Evaluation through player ratings, interviews, and sentiment analysis shows that emotional arc integration significantly enhances engagement, narrative coherence, and emotional impact. These results highlight the potential of emotionally structured procedural generation for advancing interactive storytelling for games.

Ahmed Khalifa, Roberto Gallotta, Matthew Barthet et al.

This paper introduces the Procedural Content Generation Benchmark for evaluating generative algorithms on different game content creation tasks. The benchmark comes with 12 game-related problems with multiple variants on each problem. Problems vary from creating levels of different kinds to creating rule sets for simple arcade games. Each problem has its own content representation, control parameters, and evaluation metrics for quality, diversity, and controllability. This benchmark is intended as a first step towards a standardized way of comparing generative algorithms. We use the benchmark to score three baseline algorithms: a random generator, an evolution strategy, and a genetic algorithm. Results show that some problems are easier to solve than others, as well as the impact the chosen objective has on quality, diversity, and controllability of the generated artifacts.