RESEARCH PAPERYear 2026

Hierarchical Latent Action Model

Hanjung Kim; Lerrel Pinto; Seon Joo Kim

Classification

View four quadrants
Major category
VLA
Architecture
Not applicable
Prediction paradigm
Not applicable
Source review status
Verified from primary sources

Category review. HiLAM learns variable-duration latent skills and then trains a language-conditioned hierarchical policy whose low-level module outputs executable actions. The deployed policy does not run a future-world model or planning loop. This is a latent-skill policy-learning method rather than a canonical general component. Reading evidence

AT A GLANCE

Contribution

A contribution summary has not been added yet.

Abstract

An abstract has not been added yet.

Affiliations

Not listed in the collection.

BibTeX