Zehua Jiang

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.

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.

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.

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.

Zehua Jiang, Fenghao Zhu, Xinquan Wang et al.

Generative models based on flow matching have emerged as a powerful paradigm for inverse problems, offering straighter trajectories and faster sampling compared to diffusion models. However, existing approaches often necessitate differentiating through unrolled paths, leading to numerical instability and prohibitive computational overhead. To address this, we propose P-Flow, a framework that stabilizes the reconstruction process by leveraging a proxy gradient to update the source point. This approach effectively circumvents the numerical instability and memory overhead of long-chain differentiation. To ensure consistency with the prior distribution, we employ a Gaussian spherical projection motivated by the concentration of measure phenomenon in high-dimensional spaces. We further provide a theoretical analysis for P-Flow based on Bayesian theory and Lipschitz continuity. Experiments across diverse restoration tasks demonstrate that P-Flow delivers competitive performance, especially under extreme degradations such as severely ill-posed conditions and high measurement noise.

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.

Sam Earle, Graham Todd, Yuchen Li et al.

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.

Shouren Wang, Zehua Jiang, Fernando Sliva et al.

Deep reinforcement learning (DRL) has effectively enhanced gameplay experiences and game design across various game genres. However, few studies on fighting game agents have focused explicitly on enhancing player enjoyment, a critical factor for both developers and players. To address this gap and establish a practical baseline for designing enjoyability-focused agents, we propose a two-tier agent (TTA) system and conducted experiments in the classic fighting game Street Fighter II. The first tier of TTA employs a task-oriented network architecture, modularized reward functions, and hybrid training to produce diverse and skilled DRL agents. In the second tier of TTA, a Large Language Model Hyper-Agent, leveraging players' playing data and feedback, dynamically selects suitable DRL opponents. In addition, we investigate and model several key factors that affect the enjoyability of the opponent. The experiments demonstrate improvements from 64. 36% to 156. 36% in the execution of advanced skills over baseline methods. The trained agents also exhibit distinct game-playing styles. Additionally, we conducted a small-scale user study, and the overall enjoyment in the player's feedback validates the effectiveness of our TTA system.

Sam Earle, Zehua Jiang, Eugene Vinitsky et al.

Procedural Content Generation via Reinforcement Learning (PCGRL) offers a method for training controllable level designer agents without the need for human datasets, using metrics that serve as proxies for level quality as rewards. Existing PCGRL research focuses on single generator agents, but are bottlenecked by the need to frequently recalculate heuristics of level quality and the agent's need to navigate around potentially large maps. By framing level generation as a multi-agent problem, we mitigate the efficiency bottleneck of single-agent PCGRL by reducing the number of reward calculations relative to the number of agent actions. We also find that multi-agent level generators are better able to generalize to out-of-distribution map shapes, which we argue is due to the generators' learning more local, modular design policies. We conclude that treating content generation as a distributed, multi-agent task is beneficial for generating functional artifacts at scale.

Meng Wang, Tian Lin, Aidi Lin et al.

Previous foundation models for fundus images were pre-trained with limited disease categories and knowledge base. Here we introduce a knowledge-rich vision-language model (RetiZero) that leverages knowledge from more than 400 fundus diseases. For RetiZero's pretraining, we compiled 341,896 fundus images paired with texts, sourced from public datasets, ophthalmic literature, and online resources, encompassing a diverse range of diseases across multiple ethnicities and countries. RetiZero exhibits remarkable performance in several downstream tasks, including zero-shot disease recognition, image-to-image retrieval, AI-assisted clinical diagnosis,few-shot fine-tuning, and internal- and cross-domain disease identification. In zero-shot scenarios, RetiZero achieves Top-5 accuracies of 0.843 for 15 diseases and 0.756 for 52 diseases. For image retrieval, it achieves Top-5 scores of 0.950 and 0.886 for the same sets, respectively. AI-assisted clinical diagnosis results show that RetiZero's Top-3 zero-shot performance surpasses the average of 19 ophthalmologists from Singapore, China, and the United States. RetiZero substantially enhances clinicians' accuracy in diagnosing fundus diseases, in particularly rare ones. These findings underscore the value of integrating the RetiZero into clinical settings, where various fundus diseases are encountered.

Matthew Siper, Sam Earle, Zehua Jiang et al.

Path of Destruction (PoD) is a self-supervised method for learning iterative generators. The core idea is to produce a training set by destroying a set of artifacts, and for each destructive step create a training instance based on the corresponding repair action. A generator trained on this dataset can then generate new artifacts by repairing from arbitrary states. The PoD method is very data-efficient in terms of original training examples and well-suited to functional artifacts composed of categorical data, such as game levels and discrete 3D structures. In this paper, we extend the Path of Destruction method to allow designer control over aspects of the generated artifacts. Controllability is introduced by adding conditional inputs to the state-action pairs that make up the repair trajectories. We test the controllable PoD method in a 2D dungeon setting, as well as in the domain of small 3D Lego cars.