Guardado en:
Detalles Bibliográficos
Autores principales: Zhang, Tianqiu, Lyu, Muyang, Zhang, Yufan, Fang, Fang, Wu, Si
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:https://arxiv.org/abs/2605.15725
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866911688090976256
author Zhang, Tianqiu
Lyu, Muyang
Zhang, Yufan
Fang, Fang
Wu, Si
author_facet Zhang, Tianqiu
Lyu, Muyang
Zhang, Yufan
Fang, Fang
Wu, Si
contents Latent Action Models (LAMs) enable the learning of world models from unlabeled video by inferring abstract actions between consecutive frames. However, LAMs face a fundamental trade-off between action abstraction and generation fidelity. Existing methods typically circumvent this issue by using two-stage training with pre-trained world models or by limiting predictions to optical flow. In this paper, we introduce DiLA, a novel Disentangled Latent Action world model that aims to resolve this trade-off via content-structure disentanglement. Our key insight is that disentanglement and latent action learning are co-evolving: the predictive bottleneck inherent in latent action learning serves as a driving force for disentanglement, compelling the model to distill spatial layouts into the structure pathway while offloading visual details to a separate content pathway for generation. This synergy yields a continuous, semantically structured latent action space without compromising generative quality. DiLA achieves superior results in video generation quality, action transfer, visual planning, and manifold interpretability. These findings establish DiLA as a unified framework that simultaneously achieves high-level action abstraction and high-fidelity generation, advancing the frontier of self-supervised world model learning.
format Preprint
id arxiv_https___arxiv_org_abs_2605_15725
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle DiLA: Disentangled Latent Action World Models
Zhang, Tianqiu
Lyu, Muyang
Zhang, Yufan
Fang, Fang
Wu, Si
Computer Vision and Pattern Recognition
Artificial Intelligence
Robotics
Latent Action Models (LAMs) enable the learning of world models from unlabeled video by inferring abstract actions between consecutive frames. However, LAMs face a fundamental trade-off between action abstraction and generation fidelity. Existing methods typically circumvent this issue by using two-stage training with pre-trained world models or by limiting predictions to optical flow. In this paper, we introduce DiLA, a novel Disentangled Latent Action world model that aims to resolve this trade-off via content-structure disentanglement. Our key insight is that disentanglement and latent action learning are co-evolving: the predictive bottleneck inherent in latent action learning serves as a driving force for disentanglement, compelling the model to distill spatial layouts into the structure pathway while offloading visual details to a separate content pathway for generation. This synergy yields a continuous, semantically structured latent action space without compromising generative quality. DiLA achieves superior results in video generation quality, action transfer, visual planning, and manifold interpretability. These findings establish DiLA as a unified framework that simultaneously achieves high-level action abstraction and high-fidelity generation, advancing the frontier of self-supervised world model learning.
title DiLA: Disentangled Latent Action World Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Robotics
url https://arxiv.org/abs/2605.15725