HYDRA: Representation Harmonized Tokenization for Multimodal Generation and Understanding

arXiv:2603.15228 · cs.CV · Submitted 2026-03-16 · Read on arXiv

cs.CV

Submitted: 2026-03-16

Updated: 2026-10-09

Code: https://github.com/kakaobrain/coyo-dataset

Terminology

• Benchmark — 2× in this paper · explained in 2 episodes
• Quantization — 1× in this paper · explained in 12 episodes
• Tokenization — 1× in this paper · explained in 2 episodes
• Unified Framework — 1× in this paper · explained in 5 episodes

Sources

• Qwen Technical Report
• HunyuanImage 3.0 Technical Report
• BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset
• 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
• The Prism Hypothesis: Harmonizing Semantic and Pixel Representations via Unified Autoencoding
• MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models
• SEED-X: Multimodal Models with Unified Multi-granularity Comprehension and Generation
• MetaMorph: Learning Universal Controllers with Transformers
• Omni-View: Unlocking How Generation Facilitates Understanding in Unified 3D Model based on Multiview images
• ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment
• Ming-UniVision: Joint Image Understanding and Generation with a Unified Continuous Tokenizer
• The Platonic Representation Hypothesis
• FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space
• SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehension
• LLaVA-OneVision: Easy Visual Task Transfer
• UniF$^2$ace: A Unified Fine-grained Face Understanding and Generation Model
• Evaluating Object Hallucination in Large Vision-Language Models
• Mogao: An Omni Foundation Model for Interleaved Multi-Modal Generation

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models