V-JEPA 2
Meta's world model for physical world understanding. Video Joint Embedding Predictive Architecture 2 (V-JEPA 2) is the first world model trained on video that achieves state-of-the-art visual understanding... V-JEPA 2 is Meta's new world model, trained on video to understand and predict the physical world.
Reactions (1)
Similar products
by description similarity
Viggle JST-2
viggle.aiJST-2: 3D scenes, characters and motion from one model
Molmo 2
allenai.orgSOTA video understanding, pointing, and tracking VLM
Lance
github.comimage/video generation and understanding in one model
RSI-Anything
github.comJEPA world models for self-improving agents
Google EmbeddingGemma 2
blog.googleAn open, lightweight multimodal embedding model