Skip to content
Agientry
← Back to Videos

UniAction Unifies Four Action Types in World Models

Loading video
Loading video

UniAction maps force, camera, hand, and robot-arm trajectories into one pretrained video model, with cardinality-calibrated attention for mixed action controls.

World ModelsUnified Action ConditioningRobot PoseHand PoseCamera ControlForce controlVideo GenerationWan2.2
Category: world-model
Author: @zizhang_li
Date: 2026-10-26T00:00:00
Duration: 15.0s
Source: https://x.com/zizhang_li/status/2114774656603590673
Reference: https://github.com/KBRASK/UniAction

Related Videos

Multics: Multi-Player World Model Scaling to 10 Players
0:30

Multics: Multi-Player World Model Scaling to 10 Players

@XihuiLiu · 224 views · 2026-10-14
World ModelsVideo Generation
Programmable World Model: Executable State Separated From Rendering
1:17

Programmable World Model: Executable State Separated From Rendering

@alayastd · 794 views · 2026-09-09
World ModelsVideo Generation
LingBot-World 2.0 Open-Sources Real-Time Interactive World Models
1:18

LingBot-World 2.0 Open-Sources Real-Time Interactive World Models

@robbyant_brain · 692 views · 2026-09-09
World ModelsVideo Generation
MUNDO: A Few Photos Become an Interactive, Simulatable World
2:22

MUNDO: A Few Photos Become an Interactive, Simulatable World

@zunnanxu · 45 views · 2026-10-23
MUNDOWorld Models
SKY-001 Builds an Engineering Flange in 2.87 Seconds
5:11

SKY-001 Builds an Engineering Flange in 2.87 Seconds

@skyfallai · 102 views · 2026-10-22
Engineering World ModelSkyfall AI
Juno: One Action-Conditioned JEPA Across the VLA Lifecycle
0:52

Juno: One Action-Conditioned JEPA Across the VLA Lifecycle

@walterzhu8 · 109 views · 2026-10-19
World ModelsJEPA

As an Amazon Associate, we earn from qualifying purchases.