Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generation
Paper • 2604.24763 • Published • 71
Zhang
Shangzhan
AI & ML interests
None yet
Recent Activity
upvoted a paper 2 months ago
Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generation upvoted a paper 5 months ago
Causal World Modeling for Robot Control upvoted a paper 6 months ago
Self-Evaluation Unlocks Any-Step Text-to-Image GenerationOrganizations
None yet