LatentGandr: Visual Exploration of Generative AI Latent Space via Local Embeddings

2026-04-21

AI Summary

Current generative models lack efficient and intuitive mechanisms for navigating and controlling high-dimensional latent spaces, particularly suffering from scalability and usability bottlenecks when manipulating multiple semantic attributes simultaneously. This work proposes LatentGandr, a visualization-driven analysis method grounded in local geometric structure: by automatically computing localized principal component analysis (PCA) within local neighborhoods of the latent space and integrating topological and curvature information, it interactively presents semantically meaningful directions through image grids. Departing from conventional global dimensionality reduction, LatentGandr enables finer-grained and more scalable control over generated content. User studies demonstrate that LatentGandr significantly outperforms the state-of-the-art GANSlider in both exploration efficiency and interactive experience.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Natural Language Processing: Generation
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Generative AI has demonstrated significant potential in creative design, enabling the rapid generation of visual content and imaginative concepts. Although deep AI models achieve effective featurization in the latent space, navigating the space remains a challenge. Current techniques, such as GANSlider and SliderSpace, use multiple sliders to generate high-dimensional vectors in generative AI's latent space. Despite applying (global) PCA to reduce the number of sliders, these approaches struggle with scalability and usability as the number of control dimensions increases. In this paper, we introduce LatentGandr, a visual analytics technique that facilitates latent space exploration by extracting locally linear dimensions from embeddings in high-dimensional latent spaces. By analyzing the topology and local curvature of the embeddings, LatentGandr automatically identifies local neighborhoods and computes their principal components using localized PCA. These local principal components are visualized as interactive image grids, allowing users to efficiently explore and control the generative process, providing an intuitive means to refine the generation of novel content and concepts. To evaluate the effectiveness of LatentGandr, we conducted a study comparing it to GANSlider, the current state-of-the-art visualization interface for generative AI models. The results offer insights into how localized exploration techniques can enhance user interaction with these models.

Problem

• latent space
• generative AI
• visual exploration
• scalability
• usability

Innovation

• latent space exploration
• local PCA
• visual analytics
• generative AI
• interactive image grids