The Latent Space: Foundation, Evolution, Mechanism, Ability, and Outlook
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915909487034368 |
|---|---|
| author | Yu, Xinlei Chen, Zhangquan He, Yongbo Fu, Tianyu Yang, Cheng Xu, Chengming Ma, Yue Hu, Xiaobin Cao, Zhe Xu, Jie Zhang, Guibin Tao, Jiale Zhang, Jiayi Ma, Siyuan Feng, Kaituo Huang, Haojie Li, Youxing Chen, Ronghao Wang, Huacan Wu, Chenglin Su, Zikun Xu, Xiaogang Yao, Kelu Wang, Kun Gao, Chen Liao, Yue Huang, Ruqi Jin, Tao Tan, Cheng Zhang, Jiangning Ren, Wenqi Fu, Yanwei Liu, Yong Wang, Yu Yue, Xiangyu Jiang, Yu-Gang Yan, Shuicheng |
| author_facet | Yu, Xinlei Chen, Zhangquan He, Yongbo Fu, Tianyu Yang, Cheng Xu, Chengming Ma, Yue Hu, Xiaobin Cao, Zhe Xu, Jie Zhang, Guibin Tao, Jiale Zhang, Jiayi Ma, Siyuan Feng, Kaituo Huang, Haojie Li, Youxing Chen, Ronghao Wang, Huacan Wu, Chenglin Su, Zikun Xu, Xiaogang Yao, Kelu Wang, Kun Gao, Chen Liao, Yue Huang, Ruqi Jin, Tao Tan, Cheng Zhang, Jiangning Ren, Wenqi Fu, Yanwei Liu, Yong Wang, Yu Yue, Xiangyu Jiang, Yu-Gang Yan, Shuicheng |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_02029 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | The Latent Space: Foundation, Evolution, Mechanism, Ability, and Outlook Yu, Xinlei Chen, Zhangquan He, Yongbo Fu, Tianyu Yang, Cheng Xu, Chengming Ma, Yue Hu, Xiaobin Cao, Zhe Xu, Jie Zhang, Guibin Tao, Jiale Zhang, Jiayi Ma, Siyuan Feng, Kaituo Huang, Haojie Li, Youxing Chen, Ronghao Wang, Huacan Wu, Chenglin Su, Zikun Xu, Xiaogang Yao, Kelu Wang, Kun Gao, Chen Liao, Yue Huang, Ruqi Jin, Tao Tan, Cheng Zhang, Jiangning Ren, Wenqi Fu, Yanwei Liu, Yong Wang, Yu Yue, Xiangyu Jiang, Yu-Gang Yan, Shuicheng Artificial Intelligence 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. |
| title | The Latent Space: Foundation, Evolution, Mechanism, Ability, and Outlook |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2604.02029 |