Causal World Modeling for Robot Control
Original research
MODEL PROFILE · World-action research
LingBot-VA
A video-action model using joint video/action attention and a pretrained video backbone.
Research organisationAnt Group
Research approachInverse dynamics
Published year2026
BackboneWan 2.2 (5B)
Overview
LingBot-VA jointly models visual rollout and robot-action prediction within a causal closed-loop system. NVIDIA’s field synthesis classifies it as an inverse-dynamics approach and describes its Mixture-of-Transformers architecture.
Source record
Core research identity and broad method were source-checked as of 2026-11-03. Independent benchmarking, commercial availability, weights and licensing have not been assessed.