NVIDIA's Long-WAM extends world-action model context to 19.2s, lifting RoboCasa success to 78.7%
arankomatsuzaki · x · 2026-10-09
NVIDIA introduces Long-WAM, a framework for scaling the context length of world-action models for robotics.
- Scaling temporal context from 0 to 19.2 seconds raises success on RoboCasa GR-1 from 63.3% to 78.7%
- The result suggests letting robots perceive longer histories of actions and observations is a direct lever for better manipulation performance
More from Embodied
- ENPIRE turns coding agents into autonomous robotics researchers, hitting 99% success on dexterous tasks — chris_j_paxton · 2026-10-09
- Odyssey-3 Tested: Frozen Backbone Trained on 20 Hours of Driving Data Transfers — testingcatalog · 2026-10-09
- IRVL Researcher to Present UHAS, HRT1 and VLA-Replica at Texas A&M Robotics Seminar — YuXiang_IRVL · 2026-10-09
- Google Gemma-Backed Hardware Hackathon at AGI House Gathers 120 Builders for On-Device AI — agihouse_org · 2026-10-09
- Odyssey launches Odyssey-3, claims SOTA world model on Physics-IQ benchmark — Scobleizer · 2026-10-09
- Figure valued at $39B: tracking Brett Adcock's promise of 100,000 robots in 4 years — annatonger · 2026-10-09