WorldFoundry: open-source unified infra for evaluating world models
OpenEnvisionLab · hf · 2026-11-10
OpenEnvisionLab and collaborators (Cambridge, Princeton, NVIDIA, KAUST and more) released WorldFoundry on arXiv, arguing visual plausibility alone doesn't make a world model: outputs may lack persistent state, physical dynamics, or causally consistent interactive responses. The project provides open-source infrastructure for reproducible inference and evaluation of generative world models across video, 3D/4D, embodied action, and interactive systems, with traceable evidence and benchmarks.
More from Research
- PSI claims proof of Aaronson-Ambainis conjecture using a research system built on Claude — burny_tech · 2026-11-10
- LingBot-Touch unifies three tactile sensor types using 4,983 hours of real recordings — CyberRobooo · 2026-11-10
- PNAS Nexus paper demonstrates experience-based training in a living tissue model — burny_tech · 2026-11-10
- NVIDIA paper: masking ~10 'intrusion heads' recovers 57% of distractor accuracy loss — rohanpaul_ai · 2026-11-10
- The Five Differential Operators of Physics, All Built From One Symbol: Nabla — burny_tech · 2026-11-10
- Control Limit Hypothesis: High Dopamine in Schizophrenia Is Compensation, Not Cause — burny_tech · 2026-11-10