AI Paper Atlas

Research map / Video / Action recognition

Learning Latent Action World Models in the Wild

Quentin Garrido, Tushar Nagarajan, Basile Terver, Nicolas Ballas, Yann LeCun, Michael Rabbat

ICML 2026 · Video › Action recognition · approach: Action anticipation and latent actions · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use sparse or noisy continuous latent actions instead of vector-quantized discrete ones (ablation)latent action world model on YoutubeTemporal-1B · in-the-wild one-step prediction errorprediction errorImproved
use continuous latent actions instead of discrete for complex actions (person entering room) (ablation)latent action world model · qualitative rollouts with IDMqualitative action captureImproved
insert artificial scene cuts to test for future-frame leakage (ablation)sparse, noisy and discrete latents · KineticsLPIPS: 0.20 -> 0.50 (sparse, high capacity)No clear effect
increase capacity of latent actions (higher capacity) (ablation)sparse latents · Kinetics cycle consistencyLPIPS increase on transfer: x1.20 -> x1.30Hurt
use higher capacity latent actions for transfer (ablation)sparse latents · Kinetics cycle consistencyLPIPS after transfer: 0.31 -> 0.24Improved
use the most or least constrained latent actions instead of balanced regularization (ablation)controller rollouts on DROID/RECON · controller LPIPSrollout qualityHurt
use discrete latents with controller on simple action spaces (ablation)DROID and RECON controller · controller rolloutsrollout qualityNo clear effect
use latent action world model with controller for planning vs action-conditioned V-JEPA 2-ACDROID planning with CEM · DROID goal reachingdistance to goal (delta xyz)No clear effect
use latent action world model with controller vs NoMaD policy baselineRECON navigation planning · RECONRPEImproved
train for more iterations (ablation)sparse and noisy latent action models · DROID planningdelta xyzImproved
increase model size (ablation)latent action world models · DROID planningdelta xyzMixed
increase training data quantity (ablation)latent action world models · DROID planningdelta xyzNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy