Skip to content
Hugging Face Daily Papers·· 7 days agoAI score35

UniSpark: Exploring Hybrid Attention for Multimodal Understanding and Generation

AI briefing

UniSpark-9B is proposed, a multimodal model with approximately 9B parameters, combining full attention with GatedDeltaNet linear attention, supporting visual question answering, text-to-image generation, instruction editing, and interleaved image-text prediction. It scores 2470.3 on MME-S and 0.83 on GenEval. The training data includes 159.7M generation samples, 213.4M understanding samples, and 624.8K interleaved sequences.

Open original arxiv.org