Salvato in:
Dettagli Bibliografici
Autori principali: Lee, Jung Min, Cho, Taehyun, Zhao, Li, Lee, Jungwoo
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2605.20223
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917512000569344
author Lee, Jung Min
Cho, Taehyun
Zhao, Li
Lee, Jungwoo
author_facet Lee, Jung Min
Cho, Taehyun
Zhao, Li
Lee, Jungwoo
contents Latent action models (LAMs) aim to learn action-like representations from unlabeled videos by compressing frame-to-frame changes. The frames of in-the-wild videos, however, contain not only the agent's own state but exogenous state such as background clutter. Since the exogenous state introduces changes unrelated to actions, it hinders reliable latent action learning. This paper investigates this problem analytically by extending a linear LAM framework to explicitly model exogenous state. Our analysis reveals two insights: (1) minimizing the standard reconstruction objective produces latent actions that encode exogenous information from future observation; and (2) learning in a representation space that focuses on endogenous components is a key to mitigating the interference of noise. We further show that previously proposed auxiliary objectives, such as action-supervision, provably encourage latent actions to be consistent across exogenous states. These findings are validated through experiments on both linear and nonlinear LAMs, providing a unified theoretical analysis of how exogenous state hinders latent action learning and why common remedies work.
format Preprint
id arxiv_https___arxiv_org_abs_2605_20223
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Why Latent Actions Fail, and How to Prevent It
Lee, Jung Min
Cho, Taehyun
Zhao, Li
Lee, Jungwoo
Computer Vision and Pattern Recognition
Latent action models (LAMs) aim to learn action-like representations from unlabeled videos by compressing frame-to-frame changes. The frames of in-the-wild videos, however, contain not only the agent's own state but exogenous state such as background clutter. Since the exogenous state introduces changes unrelated to actions, it hinders reliable latent action learning. This paper investigates this problem analytically by extending a linear LAM framework to explicitly model exogenous state. Our analysis reveals two insights: (1) minimizing the standard reconstruction objective produces latent actions that encode exogenous information from future observation; and (2) learning in a representation space that focuses on endogenous components is a key to mitigating the interference of noise. We further show that previously proposed auxiliary objectives, such as action-supervision, provably encourage latent actions to be consistent across exogenous states. These findings are validated through experiments on both linear and nonlinear LAMs, providing a unified theoretical analysis of how exogenous state hinders latent action learning and why common remedies work.
title Why Latent Actions Fail, and How to Prevent It
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.20223