arXiv subjects

Mahsa Bazzaz

Publications and source records attributed to Mahsa Bazzaz.

9 recordsLinked to original sources

arXiv2026

Player Perceptions of Generative AI in Games: A Steam Review Analysis

The rapid adoption of generative AI in game development has created large discussions among players, yet little empirical work has examined how players actually perceive AI-generated content. Employing quantitative methods, we study the adoption of generative AI in games in the Steam marketplace, using procedural content generation (PCG) as a baseline of a generative technology that was successfully integrated into games over several decades. Furthermore, using qualitative methods, we study player reception of generative AI by analyzing 508,192 English-language reviews. We found that games disclosing generative AI use receive lower recommendation rates and more negative overall sentiment than PCG games. Thematic analysis of 600 reviews shows that players perceive the use of generative AI in games as low developer investment in the game. Drawing on human-centered AI frameworks, we argue that successful generative AI adoption requires deploying generative AI for what players need, not for what makes development cheaper.

cs.HC↗

arXiv2026

Geometric Signatures of Reasoning: A Spectral Perspective on Task Hardness

Chain-of-thought (CoT) reasoning enables large language models (LLMs) to solve complex problems by generating intermediate reasoning steps. While much attention has been paid to the length and content of these reasoning chains, far less is known about their internal geometry. We study the \emph{geometry} of CoT trajectories in the hidden state space of transformer models, formalizing each reasoning chain as a discrete curve in $\mathbb{R}^d$ and characterizing it through spectral, positional, and kinematic geometric functionals. We introduce the effective dimension $d_\rho$ as a measure of trajectory complexity and show theoretically that trajectories with flatter eigenvalue spectra correspond to harder tasks, as they explore more of the hidden dimensions. Lastly, we explore how kinematic features of the trajectory, mean position, positional dispersion, initial and current hidden states, mean velocity, mean speed, and speed dispersion, can be used to predict solution correctness before generation is complete, and may inform future early-stopping strategies. Experimentally, on mathematical reasoning problems from the MATH500 dataset, $d_\rho$ achieves $0.93$ AUC in distinguishing easy from hard problems, while kinematic features potentially can predict correctness from only the first $20\%$ of generated tokens. These correctness signatures transfer across questions of varying difficulty, establishing that the shape of a model's internal reasoning trajectory is a principled window into both task hardness and solution quality.

cs.LG↗

arXiv2026

Playing the Imitation Game: How Perceived Generated Content Shapes Player Experience

With the fast progress of generative AI in recent years, more games are integrating generated content, raising questions regarding how players perceive and respond to this content. To investigate, we ran a mixed-method survey on the games Super Mario Bros. and Sokoban, comparing procedurally generated levels and levels designed by humans to explore how perceptions of the creator relate to players' overall experience of gameplay. Players could not reliably identify the level's creator, yet their experiences were strongly linked to their beliefs about that creator rather than the actual truth. Levels believed to be human-made were rated as more fun and aesthetically pleasing. In contrast, those believed to be AI-generated were rated as more frustrating and challenging. This negative bias appeared spontaneously without knowing the levels' creator and often was based on unreliable cues of "human-likeness." Our results underscore the importance of understanding perception biases when integrating generative systems into games.

cs.HC↗

arXiv2025

Analysis of Robustness of a Large Game Corpus

Procedural content generation via machine learning (PCGML) in games involves using machine learning techniques to create game content such as maps and levels. 2D tile-based game levels have consistently served as a standard dataset for PCGML because they are a simplified version of game levels while maintaining the specific constraints typical of games, such as being solvable. In this work, we highlight the unique characteristics of game levels, including their structured discrete data nature, the local and global constraints inherent in the games, and the sensitivity of the game levels to small changes in input. We define the robustness of data as a measure of sensitivity to small changes in input that cause a change in output, and we use this measure to analyze and compare these levels to state-of-the-art machine learning datasets, showcasing the subtle differences in their nature. We also constructed a large dataset from four games inspired by popular classic tile-based games that showcase these characteristics and address the challenge of sparse data in PCGML by providing a significantly larger dataset than those currently available.

cs.LG↗

arXiv2025

Level Generation with Constrained Expressive Range

Expressive range analysis is a visualization-based technique used to evaluate the performance of generative models, particularly in game level generation. It typically employs two quantifiable metrics to position generated artifacts on a 2D plot, offering insight into how content is distributed within a defined metric space. In this work, we use the expressive range of a generator as the conceptual space of possible creations. Inspired by the quality diversity paradigm, we explore this space to generate levels. To do so, we use a constraint-based generator that systematically traverses and generates levels in this space. To train the constraint-based generator we use different tile patterns to learn from the initial example levels. We analyze how different patterns influence the exploration of the expressive range. Specifically, we compare the exploration process based on time, the number of successful and failed sample generations, and the overall interestingness of the generated levels. Unlike typical quality diversity approaches that rely on random generation and hope to get good coverage of the expressive range, this approach systematically traverses the grid ensuring more coverage. This helps create unique and interesting game levels while also improving our understanding of the generator's strengths and limitations.

cs.LG↗

arXiv2025

Analysis of Uncertainty in Procedural Maps in Slay the Spire

This work investigates the role of uncertainty in Slay the Spire using an information-theoretic framework. Focusing on the entropy of game paths (which are based on procedurally-generated maps) we analyze how randomness influences player decision-making and success. By examining a dataset of 20,000 game runs, we quantify the entropy of paths taken by players and relate it with their outcomes and skill levels. The results show that victorious runs are associated with higher normalized entropy, suggesting more risk-taking. Additionally, higher-skill players tend to exhibit distinct patterns of risk-taking behavior in later game stages.

cs.IT↗

arXiv2024

Guided Game Level Repair via Explainable AI

Procedurally generated levels created by machine learning models can be unsolvable without further editing. Various methods have been developed to automatically repair these levels by enforcing hard constraints during the post-processing step. However, as levels increase in size, these constraint-based repairs become increasingly slow. This paper proposes using explainability methods to identify specific regions of a level that contribute to its unsolvability. By assigning higher weights to these regions, constraint-based solvers can prioritize these problematic areas, enabling more efficient repairs. Our results, tested across three games, demonstrate that this approach can help to repair procedurally generated levels faster.

cs.AI↗

arXiv2024

Controllable Game Level Generation: Assessing the Effect of Negative Examples in GAN Models

Generative Adversarial Networks (GANs) are unsupervised models designed to learn and replicate a target distribution. The vanilla versions of these models can be extended to more controllable models. Conditional Generative Adversarial Networks (CGANs) extend vanilla GANs by conditioning both the generator and discriminator on some additional information (labels). Controllable models based on complementary learning, such as Rumi-GAN, have been introduced. Rumi-GANs leverage negative examples to enhance the generator's ability to learn positive examples. We evaluate the performance of two controllable GAN variants, CGAN and Rumi-GAN, in generating game levels targeting specific constraints of interest: playability and controllability. This evaluation is conducted under two scenarios: with and without the inclusion of negative examples. The goal is to determine whether incorporating negative examples helps the GAN models avoid generating undesirable outputs. Our findings highlight the strengths and weaknesses of each method in enforcing the generation of specific conditions when generating outputs based on given positive and negative examples.

cs.LG↗

arXiv2023

Active Learning for Classifying 2D Grid-Based Level Completability

Determining the completability of levels generated by procedural generators such as machine learning models can be challenging, as it can involve the use of solver agents that often require a significant amount of time to analyze and solve levels. Active learning is not yet widely adopted in game evaluations, although it has been used successfully in natural language processing, image and speech recognition, and computer vision, where the availability of labeled data is limited or expensive. In this paper, we propose the use of active learning for learning level completability classification. Through an active learning approach, we train deep-learning models to classify the completability of generated levels for Super Mario Bros., Kid Icarus, and a Zelda-like game. We compare active learning for querying levels to label with completability against random queries. Our results show using an active learning approach to label levels results in better classifier performance with the same amount of labeled data.

cs.LG↗