Julian Togelius
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Julian Togelius has a total of 324 co-authors. Here are the top co-authors by number of shared publications:
- Georgios N. Yannakakis
Institute of Digital Games, University of Malta, Msida, Malta
29 shared publications - Ahmed Khalifa
Game Innovation Lab New York University Brooklyn USA
12 shared publications - Ahmed Khalifa
Institute of Digital Games, University of Malta, Msida, Malta
9 shared publications - Sam Earle
New York University
9 shared publications - Sebastian Risi
IT University of Copenhagen, Copenhagen, Denmark
9 shared publications - Antonios Liapis
University of Malta
8 shared publications
Latest Publications
- Open Access
- 13 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 477-480
Continuous Program Search
Genetic Programming yields highly interpretable programs, but naïve syntactic mutations can induce large and unpredictable behavioral shifts. Still, evolution shines when small changes indeed result in incremental movements on smooth fitness landscapes. We exploit this potential in the context of automated financial trading, approaching the problem of locality by treating it as an operator design problem. To this end, we learn a continuous program space where latent distance has behaviorally semantic meaning, then design mutation operators that exploit this structure without changing the evolutionary optimizer. Under identical (μ + λ) evolution strategies and fixed evaluation budgets across five tradable assets, the learned mutation operator discovers strong strategies using an order of magnitude fewer evaluations and achieves the most profitable out-of-sample Sharpe ratio. Although isotropic mutation occasionally attains higher peak performance, geometry-compiled mutation yields faster, more reliable progress, demonstrating that semantically-aligned mutation can substantially improve search efficiency without modifying the underlying evolutionary algorithm.
- 13 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 477-480
Open AccessContinuous Program Search
Genetic Programming yields highly interpretable programs, but naïve syntactic mutations can induce large and unpredictable behavioral shifts. Still, evolution shines when small changes indeed result in incremental movements on smooth fitness landscapes. We exploit this potential in the context of automated financial trading, approaching the problem of locality by treating it as an operator design problem. To this end, we learn a continuous program space where latent distance has behaviorally semantic meaning, then design mutation operators that exploit this structure without changing the evolutionary optimizer. Under identical (μ + λ) evolution strategies and fixed evaluation budgets across five tradable assets, the learned mutation operator discovers strong strategies using an order of magnitude fewer evaluations and achieves the most profitable out-of-sample Sharpe ratio. Although isotropic mutation occasionally attains higher peak performance, geometry-compiled mutation yields faster, more reliable progress, demonstrating that semantically-aligned mutation can substantially improve search efficiency without modifying the underlying evolutionary algorithm.
- Open Access
- 13 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1122-1130
BiasQD-DT: Creating Interpretable Trade-Offs in Classifier Bias with Quality-Diversity Optimization via Decision Trees
The implementation of Artificial Intelligence (AI) in sensitive domains requires that models are fair and clearly understandable for stakeholders to trust in the decisions made. While Explainable AI (XAI) methods approximate the behavior of models, Interpretable AI (IAI) focuses on inherently transparent representations such as decision trees. However, interpretability alone is not enough, as models may encode biases from different sources, including biases present in data, leading to unfair outcomes. This introduces a fundamental trade-off between predictive performance and fairness. In this paper, we propose a novel framework that integrates evolved decision trees generated from random seed-based vectors and Quality Diversity (QD) optimization to jointly address decision interpretability, predictive performance, and fairness. Using evolutionary search, the method generates a repertoire of high-performing, interpretable models that span different regions of the fairness-accuracy space. This enables a systematic characterization of trade-offs and provides decision-makers with multiple transparent alternatives. Experimental results show that the proposed approach effectively discovers diverse decision tree models with competitive overall accuracy while achieving improved fairness across standard benchmarks.
- 13 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1122-1130
Open AccessBiasQD-DT: Creating Interpretable Trade-Offs in Classifier Bias with Quality-Diversity Optimization via Decision Trees
The implementation of Artificial Intelligence (AI) in sensitive domains requires that models are fair and clearly understandable for stakeholders to trust in the decisions made. While Explainable AI (XAI) methods approximate the behavior of models, Interpretable AI (IAI) focuses on inherently transparent representations such as decision trees. However, interpretability alone is not enough, as models may encode biases from different sources, including biases present in data, leading to unfair outcomes. This introduces a fundamental trade-off between predictive performance and fairness. In this paper, we propose a novel framework that integrates evolved decision trees generated from random seed-based vectors and Quality Diversity (QD) optimization to jointly address decision interpretability, predictive performance, and fairness. Using evolutionary search, the method generates a repertoire of high-performing, interpretable models that span different regions of the fairness-accuracy space. This enables a systematic characterization of trade-offs and provides decision-makers with multiple transparent alternatives. Experimental results show that the proposed approach effectively discovers diverse decision tree models with competitive overall accuracy while achieving improved fairness across standard benchmarks.
- Open Access
- 10 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 310-318
Mortar: Evolving Mechanics for Automatic Game Design
We present Mortar, a system for autonomously evolving game mechanics for automatic game design. Game mechanics define the rules and interactions that govern gameplay, and designing them manually is a time-consuming and expert-driven process. Mortar combines a quality-diversity algorithm with a large language model to explore a diverse set of mechanics, which are evaluated by synthesising complete games that incorporate both evolved mechanics and those drawn from an archive. The mechanics are evaluated by composing complete games through a tree search procedure, where the resulting games are evaluated by their ability to preserve a skill-based ordering over players—that is, whether stronger players consistently outperform weaker ones. We assess the mechanics based on their contribution towards the skill-based ordering score in the game. We demonstrate that Mortar produces games that appear diverse and playable, and mechanics that contribute more towards the skill-based ordering score in the game. We perform ablation studies to assess the role of each system component and a user study to evaluate the games based on human feedback.
- 10 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 310-318
Open AccessMortar: Evolving Mechanics for Automatic Game Design
We present Mortar, a system for autonomously evolving game mechanics for automatic game design. Game mechanics define the rules and interactions that govern gameplay, and designing them manually is a time-consuming and expert-driven process. Mortar combines a quality-diversity algorithm with a large language model to explore a diverse set of mechanics, which are evaluated by synthesising complete games that incorporate both evolved mechanics and those drawn from an archive. The mechanics are evaluated by composing complete games through a tree search procedure, where the resulting games are evaluated by their ability to preserve a skill-based ordering over players—that is, whether stronger players consistently outperform weaker ones. We assess the mechanics based on their contribution towards the skill-based ordering score in the game. We demonstrate that Mortar produces games that appear diverse and playable, and mechanics that contribute more towards the skill-based ordering score in the game. We perform ablation studies to assess the role of each system component and a user study to evaluate the games based on human feedback.
- Open Access
- 10 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 727-734
ProFiT: Program Search for Financial Trading
- 10 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 727-734
Open AccessProFiT: Program Search for Financial Trading
- Open Access
- 10 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 106-114
In Search of the Ingredients of Open-Endedness: Replicating Picbreeder with Large Vision-Language Models
We are in the midst of large-scale industrial and academic efforts to automate the processes of scientific, technological and creative production through AI-driven assistants. Historically, a fundamental property of these processes in their human form has been their open-endedness: their capacity for generating a seemingly endless supply of novel and meaningful new forms. Do artificial agents have any capacity for such fruitful unguided discovery? To answer this question, we turn to Picbreeder, the canonical exemplar of human-driven open-ended search, in which users collaboratively generated a diverse library of images through interactive evolution of small neural networks. We replicate Picbreeder, replacing human users with frontier Vision Language Models (VLMs). We observe clear qualitative differences between the output of our system and the historical human baseline, and attempt to characterize them using metrics of phylogenetic complexity and visual and semantic salience and novelty. In an effort to identify some of the causal factors contributing these differences, we study the addition of exploratory noise to the agents' selection process, of behavioral diversity between agents, and of narrative momentum in the form of memory of past actions. We make our code available at https://github.com/smearle/picbreeder-vlm.
- 10 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 106-114
Open AccessIn Search of the Ingredients of Open-Endedness: Replicating Picbreeder with Large Vision-Language Models
We are in the midst of large-scale industrial and academic efforts to automate the processes of scientific, technological and creative production through AI-driven assistants. Historically, a fundamental property of these processes in their human form has been their open-endedness: their capacity for generating a seemingly endless supply of novel and meaningful new forms. Do artificial agents have any capacity for such fruitful unguided discovery? To answer this question, we turn to Picbreeder, the canonical exemplar of human-driven open-ended search, in which users collaboratively generated a diverse library of images through interactive evolution of small neural networks. We replicate Picbreeder, replacing human users with frontier Vision Language Models (VLMs). We observe clear qualitative differences between the output of our system and the historical human baseline, and attempt to characterize them using metrics of phylogenetic complexity and visual and semantic salience and novelty. In an effort to identify some of the causal factors contributing these differences, we study the addition of exploratory noise to the agents' selection process, of behavioral diversity between agents, and of narrative momentum in the form of memory of past actions. We make our code available at https://github.com/smearle/picbreeder-vlm.
- 20 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Games
- Vol. PP (24751502) , 1-14
PCGRLLM: Large Language Model-Driven Reward Design for Procedural Content Generation Reinforcement Learning
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 across various reasoning-based prompting methods. Our experiments provide insightful evaluations that demonstrate the capabilities of LLMs essential for content generation tasks. The results demonstrate a substantial performance improvement over the previous structure, achieving performance comparable to that of humans. Our work demonstrates the potential to reduce human dependency in game AI development, while supporting and enhancing creative processes.
- 20 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Games
- Vol. PP (24751502) , 1-14
PCGRLLM: Large Language Model-Driven Reward Design for Procedural Content Generation Reinforcement Learning
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 across various reasoning-based prompting methods. Our experiments provide insightful evaluations that demonstrate the capabilities of LLMs essential for content generation tasks. The results demonstrate a substantial performance improvement over the previous structure, achieving performance comparable to that of humans. Our work demonstrates the potential to reduce human dependency in game AI development, while supporting and enhancing creative processes.
- Open Access
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
The Garden of Forking Paths: Narrative Arc-Conditioned Gameplay Planning
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessThe Garden of Forking Paths: Narrative Arc-Conditioned Gameplay Planning
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 162974-162994
Exploring the Relationship Between Training and Test-Time Augmentation in Image Classification
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 162974-162994
Open AccessExploring the Relationship Between Training and Test-Time Augmentation in Image Classification
- Open Access
Mortar: Evolving Mechanics for Automatic Game Design
We present Mortar, a system for autonomously evolving game mechanics for automatic game design. Game mechanics define the rules and interactions that govern gameplay, and designing them manually is a time-consuming and expert-driven process. Mortar combines a quality-diversity algorithm with a large language model to explore a diverse set of mechanics, which are evaluated by synthesising complete games that incorporate both evolved mechanics and those drawn from an archive. The mechanics are evaluated by composing complete games through a tree search procedure, where the resulting games are evaluated by their ability to preserve a skill-based ordering over players -- that is, whether stronger players consistently outperform weaker ones. We assess the mechanics based on their contribution towards the skill-based ordering score in the game. We demonstrate that Mortar produces games that appear diverse and playable, and mechanics that contribute more towards the skill-based ordering score in the game. We perform ablation studies to assess the role of each system component and a user study to evaluate the games based on human feedback.
Open AccessMortar: Evolving Mechanics for Automatic Game Design
We present Mortar, a system for autonomously evolving game mechanics for automatic game design. Game mechanics define the rules and interactions that govern gameplay, and designing them manually is a time-consuming and expert-driven process. Mortar combines a quality-diversity algorithm with a large language model to explore a diverse set of mechanics, which are evaluated by synthesising complete games that incorporate both evolved mechanics and those drawn from an archive. The mechanics are evaluated by composing complete games through a tree search procedure, where the resulting games are evaluated by their ability to preserve a skill-based ordering over players -- that is, whether stronger players consistently outperform weaker ones. We assess the mechanics based on their contribution towards the skill-based ordering score in the game. We demonstrate that Mortar produces games that appear diverse and playable, and mechanics that contribute more towards the skill-based ordering score in the game. We perform ablation studies to assess the role of each system component and a user study to evaluate the games based on human feedback.
- Open Access
PuzzleJAX: A Benchmark for Reasoning and Learning
We introduce PuzzleJAX, a GPU-accelerated puzzle game engine and description language designed to support rapid benchmarking of tree search, reinforcement learning, and LLM reasoning abilities. Unlike existing GPU-accelerated learning environments that provide hard-coded implementations of fixed sets of games, PuzzleJAX allows dynamic compilation of any game expressible in its domain-specific language (DSL). This DSL follows PuzzleScript, which is a popular and accessible online game engine for designing puzzle games. In this paper, we validate in PuzzleJAX several hundred of the thousands of games designed in PuzzleScript by both professional designers and casual creators since its release in 2013, thereby demonstrating PuzzleJAX's coverage of an expansive, expressive, and human-relevant space of tasks. By analyzing the performance of search, learning, and language models on these games, we show that PuzzleJAX can naturally express tasks that are both simple and intuitive to understand, yet often deeply challenging to master, requiring a combination of control, planning, and high-level insight.
Open AccessPuzzleJAX: A Benchmark for Reasoning and Learning
We introduce PuzzleJAX, a GPU-accelerated puzzle game engine and description language designed to support rapid benchmarking of tree search, reinforcement learning, and LLM reasoning abilities. Unlike existing GPU-accelerated learning environments that provide hard-coded implementations of fixed sets of games, PuzzleJAX allows dynamic compilation of any game expressible in its domain-specific language (DSL). This DSL follows PuzzleScript, which is a popular and accessible online game engine for designing puzzle games. In this paper, we validate in PuzzleJAX several hundred of the thousands of games designed in PuzzleScript by both professional designers and casual creators since its release in 2013, thereby demonstrating PuzzleJAX's coverage of an expansive, expressive, and human-relevant space of tasks. By analyzing the performance of search, learning, and language models on these games, we show that PuzzleJAX can naturally express tasks that are both simple and intuitive to understand, yet often deeply challenging to master, requiring a combination of control, planning, and high-level insight.