Paper 2604.02029
The Latent Space: Foundation, Evolution, Mechanism, Ability, and Outlook
- Published
- Apr 2026
- Research lab
- Independent
- Citations
- 30
- GitHub
- 956 stars
01 In brief
Summary
This survey provides a unified overview of latent space in language-based models, arguing that it is emerging as a native computational substrate beyond explicit token-level processing.
It addresses the fragmentation in the field by organizing research into five sequential perspectives: Foundation, Evolution, Mechanism, Ability, and Outlook.
The survey defines latent space as the continuous, learned representational space in models, contrasting it with explicit verbal space and the latent spaces of generative visual models.
It traces the field's evolution through four stages: Prototype, Formation, Expansion, and Outbreak.
A two-dimensional taxonomy is introduced, classifying methods by Mechanism (Architecture, Representation, Computation, Optimization) and Ability (Reasoning, Planning, Modeling, Perception, Memory, Collaboration, Embodiment).
The survey concludes by discussing open challenges in evaluability, controllability, and interpretability, and outlines future directions for theory, multimodal integration, and downstream tasks, positioning latent space as a potential foundational principle for next-generation intelligence.
The authors also maintain a GitHub repository for community contributions and updates.
The survey is authored by a large collaborative team from multiple institutions and is available on arXiv (2604.02029v2).
02 From the paper
Abstract
Latent space is rapidly emerging as a native substrate for language-based models. While modern systems are still commonly understood through explicit token-level generation, an increasing body of work shows that many critical internal processes are more naturally carried out in continuous latent space than in human-readable verbal traces. This shift is driven by the structural limitations of explicit-space computation, including linguistic redundancy, discretization bottlenecks, sequential inefficiency, and semantic loss. This survey aims to provide a unified and up-to-date landscape of latent space in language-based models. We organize the survey into five sequential perspectives: Foundation, Evolution, Mechanism, Ability, and Outlook. We begin by delineating the scope of latent space, distinguishing it from explicit or verbal space and from the latent spaces commonly studied in generative visual models. We then trace the field's evolution from early exploratory efforts to the current large-scale expansion. To organize the technical landscape, we examine existing work through the complementary lenses of mechanism and ability. From the perspective of Mechanism, we identify four major lines of development: Architecture, Representation, Computation, and Optimization. From the perspective of Ability, we show how latent space supports a broad capability spectrum spanning Reasoning, Planning, Modeling, Perception, Memory, Collaboration, and Embodiment. Beyond consolidation, we discuss the key open challenges, and outline promising directions for future research. We hope this survey serves not only as a reference for existing work, but also as a foundation for understanding latent space as a general computational and systems paradigm for next-generation intelligence.