mlsanityData to 14 Oct 2026

MMaDA: Multimodal Large Diffusion Language Models

arXiv 2505.15809 ↗First posted 21 May 2025cs.CVcited by 250

On the map: Diffusion models → Image generation, text-to-image and editing

Abstract (arXiv)

We introduce MMaDA, a novel class of multimodal diffusion foundation models designed to achieve superior performance across diverse domains such as textual reasoning, multimodal understanding, and text-to-image generation. The approach is distinguished by three key innovations: (i) MMaDA adopts a unified diffusion architecture with a shared probabilistic formulation and a modality-agnostic design, eliminating the need for modality-specific components. This architecture ensures seamless integration and processing across different data types. (ii) We implement a mixed long chain-of-thought (CoT) fine-tuning strategy that curates a unified CoT format across modalities. By aligning reasoning processes between textual and visual domains, this strategy facilitates cold-start training for the final reinforcement learning (RL) stage, thereby enhancing the model's ability to handle complex tasks from the outset. (iii) We propose UniGRPO, a unified policy-gradient-based RL algorithm specifically tailored for diffusion foundation models. Utilizing diversified reward modeling, UniGRPO unifies post-training across both reasoning and generation tasks, ensuring consistent performance improvements. Experimental results demonstrate that MMaDA-8B exhibits strong generalization capabilities as a unified multimodal foundation model. It surpasses powerful models like LLaMA-3-7B and Qwen2-7B in textual reasoning, outperforms Show-o and SEED-X in multimodal understanding, and excels over SDXL and Janus in text-to-image generation. These achievements highlight MMaDA's effectiveness in bridging the gap between pretraining and post-training within unified diffusion architectures, providing a comprehensive framework for future research and development. We open-source our code and trained models at: https://github.com/Gen-Verse/MMaDA

Key facts. Cited by 250 papers in the map (data through 14 Oct 2026). Placed in Image generation, text-to-image and editing, one of 14 directions in Diffusion models. Builds on 89 of its 109 references that are in the map. First posted on arXiv 21 May 2025.

Builds on

89 of 109 references.

Closest ancestors · 9

Same direction, newest first

  1. 2024IterComp: Iterative Composition-Aware Feedback Learning from Model Gallery for Text-to-Image Generation9 Oct
  2. 2024Emu3: Next-Token Prediction is All You Need27 Sep
  3. 2024VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation6 Sep
  4. 2024Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction3 Apr
  5. 2024Mastering Text-to-Image Diffusion: Recaptioning, Planning, and Generating with Multimodal LLMs22 Jan
  6. 2023GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignment17 Oct
  7. 2023RAPHAEL: Text-to-Image Generation via Large Mixture of Diffusion Paths29 May
  8. 2022Hierarchical Text-Conditional Image Generation with CLIP Latents13 Apr
  9. 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models20 Dec

Other references · 80

  1. 2017Attention Is All You Needcited by 60,361
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2020Language Models are Few-Shot Learnerscited by 32,419
  4. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  5. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  6. 2023GPT-4 Technical Reportcited by 18,396
  7. 2017Proximal Policy Optimization Algorithmscited by 15,215
  8. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,599
  9. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  10. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  11. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  12. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
38 more
  1. 2023Segment Anythingcited by 8,234
  2. 2023Visual Instruction Tuningcited by 7,869
  3. 2023Improved Baselines with Visual Instruction Tuningcited by 4,418
  4. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,356
  5. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  6. 2023Qwen Technical Reportcited by 3,344
  7. 2020Taming Transformers for High-Resolution Image Synthesiscited by 3,253
  8. 2023InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuningcited by 2,876
  9. 2023MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Modelscited by 2,711
  10. 2015High-Dimensional Continuous Control Using Generalized Advantage Estimationcited by 2,026
  11. 2021Structured Denoising Diffusion Models in Discrete State-Spacescited by 1,719
  12. 2016CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoningcited by 1,457
  13. 2025s1: Simple test-time scalingcited by 1,341
  14. 2021Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Conceptscited by 1,216
  15. 2022MaskGIT: Masked Generative Image Transformercited by 1,099
  16. 2023ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generationcited by 1,045
  17. 2024Chameleon: Mixed-Modal Early-Fusion Foundation Modelscited by 927
  18. 2025Large Language Diffusion Modelscited by 912
  19. 2023A Survey on Multimodal Large Language Modelscited by 909
  20. 2023PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesiscited by 864
  21. 2025Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scalingcited by 812
  22. 2024Autoregressive Model Beats Diffusion: Llama for Scalable Image Generationcited by 777
  23. 2021Vector Quantized Diffusion Model for Text-to-Image Synthesiscited by 774
  24. 2018Image Transformercited by 772
  25. 2024Show-o: One Single Transformer to Unify Multimodal Understanding and Generationcited by 752
  26. 2023Muse: Text-To-Image Generation via Masked Generative Transformerscited by 648
  27. 2023Language Model Beats Diffusion -- Tokenizer is Key to Visual Generationcited by 640
  28. 2023NExT-GPT: Any-to-Any Multimodal LLMcited by 615
  29. 2021Argmax Flows and Multinomial Diffusion: Learning Categorical Distributionscited by 574
  30. 2023mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with Modality Collaborationcited by 547
  31. 2025LIMO: Less is More for Reasoningcited by 498
  32. 2023VideoPoet: A Large Language Model for Zero-Shot Video Generationcited by 492
  33. 2024Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Modelcited by 483
  34. 2024Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generationcited by 457
  35. 2023Generative Multimodal Models are In-Context Learnerscited by 446
  36. 2024Hallucination of Multimodal Large Language Models: A Surveycited by 388
  37. 2024SEED-X: Multimodal Models with Unified Multi-granularity Comprehension and Generationcited by 331
  38. 2023DreamLLM: Synergistic Multimodal Comprehension and Creationcited by 297

50 most cited of 80 All 80 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 110; 2026: 140.

250 papers.

Sort:

By area:

  1. 2025LLaDA 1.5: Variance-Reduced Preference Optimization for Large Language Diffusion Modelscited by 285
  2. 2025Motus: A Unified Latent Action World Modelcited by 251
  3. 2025Show-o2: Improved Native Unified Multimodal Modelscited by 250
  4. 2025DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generationcited by 187
  5. 2025A Survey of Reinforcement Learning for Large Reasoning Modelscited by 180
  6. 2025A Survey of Context Engineering for Large Language Modelscited by 126
  7. 2025Discrete Diffusion VLA: Bringing Discrete Diffusion to Action Decoding in Vision-Language-Action Policiescited by 90
  8. 2025Lumina-DiMOO: An Omni Diffusion Large Language Model for Multi-Modal Generation and Understandingcited by 88
  9. 2025A Survey on Latent Reasoningcited by 81
  10. 2025A Survey on Diffusion Language Modelscited by 78
  11. 2025wd1: Weighted Policy Optimization for Reasoning in Diffusion Language Modelscited by 71
  12. 2025Sparse-dLLM: Accelerating Diffusion LLMs with Dynamic Cache Evictioncited by 65
38 more
  1. 2025dParallel: Learnable Parallel Decoding for dLLMscited by 65
  2. 2025Discrete Diffusion in Large Language and Multimodal Models: A Surveycited by 61
  3. 2025LLaDA-MoE: A Sparse MoE Diffusion Language Modelcited by 57
  4. 2025Co-Evolving LLM Coder and Unit Tester via Reinforcement Learningcited by 56
  5. 2025Muddit: Liberating Generation Beyond Text-to-Image with a Unified Discrete Diffusion Modelcited by 55
  6. 2025Wide-In, Narrow-Out: Revokable Decoding for Efficient and Effective DLLMscited by 46
  7. 2026ELF: Embedded Language Flowscited by 45
  8. 2025ParallelBench: Understanding the Trade-offs of Parallel Decoding in Diffusion LLMscited by 43
  9. 2025SPG: Sandwiched Policy Gradient for Masked Diffusion Language Modelscited by 41
  10. 2025LongLLaDA: Unlocking Long Context Capabilities in Diffusion LLMscited by 40
  11. 2025Unified Diffusion VLA: Vision-Language-Action Model via Joint Discrete Denoising Diffusion Processcited by 39
  12. 2026d3LLM: Ultra-Fast Diffusion LLM using Pseudo-Trajectory Distillationcited by 37
  13. 2025The Devil behind the mask: An emergent safety vulnerability of Diffusion LLMscited by 34
  14. 2025Continuously Augmented Discrete Diffusion model for Categorical Generative Modelingcited by 34
  15. 2025Principled RL for Diffusion LLMs Emerges from a Sequence-Level Perspectivecited by 34
  16. 2025Time Is a Feature: Exploiting Temporal Dynamics in Diffusion Language Modelscited by 33
  17. 2025Efficient-DLM: From Autoregressive to Diffusion Language Models, and Beyond in Speedcited by 33
  18. 2025Multimodal Chain of Continuous Thought for Latent-Space Reasoning in Vision-Language Modelscited by 32
  19. 2025Attention Is All You Need for KV Cache in Diffusion LLMscited by 32
  20. 2025AdaBlock-dLLM: Semantic-Aware Diffusion LLM Inference via Adaptive Block Sizecited by 31
  21. 2025LLaDA-VLA: Vision Language Diffusion Action Modelscited by 30
  22. 2025dVLA: Diffusion Vision-Language-Action Model with Multimodal Chain-of-Thoughtcited by 30
  23. 2025Self Speculative Decoding for Diffusion Large Language Modelscited by 30
  24. 2026InternVL-U: Democratizing Unified Multimodal Models for Understanding, Reasoning, Generation and Editingcited by 28
  25. 2025Jailbreaking Large Language Diffusion Models: Revealing Hidden Safety Flaws in Diffusion-Based Text Generationcited by 26
  26. 2025UniGenBench++: A Unified Semantic Evaluation Benchmark for Text-to-Image Generationcited by 25
  27. 2025Encoder-Decoder Diffusion Language Models for Efficient Training and Inferencecited by 24
  28. 2026Nemotron-Labs-Diffusion: A Tri-Mode Language Model Unifying Autoregressive, Diffusion, and Self-Speculation Decodingcited by 24
  29. 2025Quantization Meets dLLMs: A Systematic Study of Post-training Quantization for Diffusion LLMscited by 23
  30. 2025Set Block Decoding is a Language Model Inference Acceleratorcited by 23
  31. 2025Lumos-1: On Autoregressive Video Generation with Discrete Diffusion from a Unified Model Perspectivecited by 22
  32. 2025Accelerating Diffusion LLM Inference via Local Determinism Propagationcited by 21
  33. 2025NExT-OMNI: Towards Any-to-Any Omnimodal Foundation Models with Discrete Flow Matchingcited by 21
  34. 2025VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoocited by 20
  35. 2025Enhancing Reasoning for Diffusion LLMs via Distribution Matching Policy Optimizationcited by 20
  36. 2025DiFFPO: Training Diffusion LLMs to Reason Fast and Furious via Reinforcement Learningcited by 19
  37. 2025Dream-VL & Dream-VLA: Open Vision-Language and Vision-Language-Action Models with Diffusion Language Model Backbonecited by 18
  38. 2026DMax: Aggressive Parallel Decoding for dLLMscited by 18

50 most cited of 250 All 250 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025I Think, Therefore I Diffuse: Enabling Multimodal In-Context Reasoning in Diffusion Models
  2. 2024Dual Diffusion for Unified Image Generation and Understanding
  3. 2025LaViDa: A Large Diffusion Language Model for Multimodal Understanding
  4. 2026DiffusionOPD: A Unified Perspective of On-Policy Distillation in Diffusion Models
  5. 2025Emu3.5: Native Multimodal Models are World Learners
  6. 2025MammothModa2: A Unified AR-Diffusion Framework for Multimodal Understanding and Generation
  7. 2026DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
  8. 2025DiffThinker: Towards Generative Multimodal Reasoning with Diffusion Models