Logit-Conditioned Diffusion Decoding for Frozen Discrete-Token VLMs
cs.CV, cs.LG
Submitted: 2026-03-11
Updated: 2026-11-09
Terminology
Sources
- Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling
- X-Omni: Reinforcement Learning Makes Discrete Autoregressive Image Generative Models Great Again
- ILLUME+: Illuminating Unified MLLM with Dual Visual Tokenization and Diffusion Refinement
- FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space
- UniFusion: Vision-Language Model as Unified Encoder in Image Generation
- Decoupled Weight Decay Regularization
- Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation
- HART: Efficient Visual Generation with Hybrid Autoregressive Transformer
- Bridging Continuous and Discrete Tokens for Autoregressive Visual Generation
- Show-o: One Single Transformer to Unify Multimodal Understanding and Generation
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