Navigation world models as group actions

Wang, Zijie, Zhang, Wei, Weiming, Zhang, Zhang, Fanqi, Tan, Xiao, Qin, Yipeng ORCID: https://orcid.org/0000-0002-1551-9126 and Li, Guanbin 2026. Navigation world models as group actions. Presented at: The Fortieth Annual Conference on Neural Information Processing Systems (NeurIPS 2026), Sydney, Australia, 6-12 December 2026. NeurIPS 2026 Proceedings. NeurIPS, Item availability restricted.

	PDF - Accepted Post-Print Version Restricted to Repository staff only Download (2MB)
	PDF (Provisional file) - Accepted Post-Print Version Download (17kB)

Abstract

Navigation video world models have achieved strong visual realism, but this does not ensure that their dynamics are truly governed by actions. In this work, we argue that action faithfulness should be understood through the compositional structure of actions, which in many embodied settings follows a group structure (e.g., SE(2) for navigation). Based on this insight, we formalize action-conditioned navigation world modeling as realizing a group action on the state space, providing a principled criterion for evaluating dynamics beyond visual quality. To operationalize this framework, we propose a unified approach that enforces identity, inverse, and composition consistency via latent-space regularization with synthesized supervision, avoiding additional data collection. We further introduce two metrics: Group-Action Consistency (GAC) and Group-Action Robustness (GAR), to evaluate structural correctness and rollout stability. Extensive experimental results show that our method consistently improves both GAC and GAR across representative video world model backbones in the navigation setting without degrading perceptual quality.

Item Type:	Conference or Workshop Item - published (Paper)
Status:	In Press
Schools:	Schools > Computational & Mathematical Sciences Schools > Computer Science & Informatics
Publisher:	NeurIPS
Date of First Compliant Deposit:	29 October 2026
Date of Acceptance:	25 September 2026
Last Modified:	29 Oct 2026 14:30
URI:	https://orca.cardiff.ac.uk/id/eprint/190410

Actions (repository staff only)

Edit Item