Skip to content
Agientry
← Back to Videos

DreamFAR: A Robot-Only World Model Trained from Scratch

Loading video
Loading video

DreamFAR trains a world model from scratch on robot data alone, enabling real-time teleoperation, minute-scale rollouts, and generalization across more than 1,000 tasks.

DreamFARWorld ModelsRobot LearningEmbodied AIGeneralist Robot Policy
Category: research
Author: @_YimingDou
Date: 2026-10-29T00:00:00
Duration: 73.542s
Source: https://x.com/_YimingDou/status/2115842800407466466
Reference: https://www.yimingdou.com/DreamFAR

Related Videos

Odyssey-3: A Foundation World Model for Physical Agents
1:37

Odyssey-3: A Foundation World Model for Physical Agents

@odysseyml · 1.2k views · 2026-09-15
World ModelsFoundation Models
Video2World Tests Whether Coding Agents Rebuild Executable Robot Worlds
0:12

Video2World Tests Whether Coding Agents Rebuild Executable Robot Worlds

@huang_biwei · 16 views · 2026-10-29
Video2WorldCoding Agent
Same VLA, three data sources: Human Archive OTS wins on task success
0:43

Same VLA, three data sources: Human Archive OTS wins on task success

@babugi28 · 157 views · 2026-10-15
VLAPre-training Data
RoboJEPA: The First Scaling Law for Multi-Embodiment World Models
1:06

RoboJEPA: The First Scaling Law for Multi-Embodiment World Models

@artemZholus · 293 views · 2026-10-10
World ModelsJEPA
JEPA-Anything: One World Model Across Six Domains
0:30

JEPA-Anything: One World Model Across Six Domains

@Charles_Y_Wu · 454 views · 2026-10-05
World ModelsEmbodied AI
Gen-HumanEgo: Open-Source Egocentric Human Data for Robot Learning
4:14

Gen-HumanEgo: Open-Source Egocentric Human Data for Robot Learning

@GenrobotAI · 1.1k views · 2026-09-23
Egocentric VisionRobot Learning

As an Amazon Associate, we earn from qualifying purchases.