UniSpark: Exploring Hybrid Attention for Multimodal Understanding and Generation
cs.CV
Submitted: 2026-11-12
Updated: 2026-11-12
Code: https://github.com/Wan-Video/Wan2.2
Project page: https://lucaria-academy.github.io/UniSpark
Terminology
Sources
- Qwen2.5-VL Technical Report
- FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space
- Chameleon: Mixed-Modal Early-Fusion Foundation Models
- SANA-Video 2.0: Hybrid Linear Attention with Attention Residuals for Efficient Video Generation
- Training Deep Nets with Sublinear Memory Cost
- Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling
- Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling
- Emerging Properties in Unified Multimodal Pretraining
- SenseNova-U1: Unifying Multimodal Understanding and Generation with NEO-unify Architecture
- MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models
- Seedream 3.0 Technical Report
- SEED-X: Multimodal Models with Unified Multi-granularity Comprehension and Generation
- GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignment
- Vision as Unified Multimodal Generation
- Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention
- ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment
- LLaDA2.0-Uni: Unifying Multimodal Understanding and Generation with Diffusion Large Language Model
- Kimi-VL Technical Report
- Kimi Linear: An Expressive, Efficient Attention Architecture
- LLaVA-OneVision: Easy Visual Task Transfer
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs