mlsanityData to 14 Oct 2026

X-Omni: Reinforcement Learning Makes Discrete Autoregressive Image Generative Models Great Again

arXiv 2507.22058 ↗First posted 29 Jul 2025cs.CVcited by 112

On the map: Deep reinforcement learning methods → RL with diffusion and flow models

Abstract (arXiv)

Numerous efforts have been made to extend the ``next token prediction'' paradigm to visual contents, aiming to create a unified approach for both image generation and understanding. Nevertheless, attempts to generate images through autoregressive modeling with discrete tokens have been plagued by issues such as low visual fidelity, distorted outputs, and failure to adhere to complex instructions when rendering intricate details. These shortcomings are likely attributed to cumulative errors during autoregressive inference or information loss incurred during the discretization process. Probably due to this challenge, recent research has increasingly shifted toward jointly training image generation with diffusion objectives and language generation with autoregressive objectives, moving away from unified modeling approaches. In this work, we demonstrate that reinforcement learning can effectively mitigate artifacts and largely enhance the generation quality of a discrete autoregressive modeling method, thereby enabling seamless integration of image and language generation. Our framework comprises a semantic image tokenizer, a unified autoregressive model for both language and images, and an offline diffusion decoder for image generation, termed X-Omni. X-Omni achieves state-of-the-art performance in image generation tasks using a 7B language model, producing images with high aesthetic quality while exhibiting strong capabilities in following instructions and rendering long texts.

Key facts. Cited by 112 papers in the map (data through 14 Oct 2026). Placed in RL with diffusion and flow models, one of 15 directions in Deep reinforcement learning methods. Builds on 79 of its 90 references that are in the map. First posted on arXiv 29 Jul 2025.

Builds on

79 of 90 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025A Simple and Effective Reinforcement Learning Method for Text-to-Image Diffusion Fine-tuning2 Mar

Other references · 78

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2023GPT-4 Technical Reportcited by 18,396
  4. 2017Proximal Policy Optimization Algorithmscited by 15,215
  5. 2024The Llama 3 Herd of Modelscited by 14,715
  6. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  7. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 13,028
  8. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  9. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  10. 2022Hierarchical Text-Conditional Image Generation with CLIP Latentscited by 6,585
  11. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 6,355
  12. 2025Qwen2.5-VL Technical Reportcited by 5,587
38 more
  1. 2024Qwen2.5 Technical Reportcited by 4,942
  2. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,703
  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. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  6. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
  7. 2023Qwen Technical Reportcited by 3,344
  8. 2024LLaVA-OneVision: Easy Visual Task Transfercited by 2,719
  9. 2023MMBench: Is Your Multi-modal Model an All-around Player?cited by 2,141
  10. 2023Evaluating Object Hallucination in Large Vision-Language Modelscited by 1,729
  11. 2020DocVQA: A Dataset for VQA on Document Imagescited by 1,317
  12. 2025SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Featurescited by 1,188
  13. 2024MiniCPM-V: A GPT-4V Level MLLM on Your Phonecited by 1,079
  14. 2023ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generationcited by 1,045
  15. 2023Training Diffusion Models with Reinforcement Learningcited by 982
  16. 2023GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignmentcited by 972
  17. 2024Chameleon: Mixed-Modal Early-Fusion Foundation Modelscited by 927
  18. 2025Emerging Properties in Unified Multimodal Pretrainingcited by 897
  19. 2023Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesiscited by 890
  20. 2025Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scalingcited by 812
  21. 2023EVA-CLIP: Improved Training Techniques for CLIP at Scalecited by 798
  22. 2023Diffusion Model Alignment Using Direct Preference Optimizationcited by 790
  23. 2024Emu3: Next-Token Prediction is All You Needcited by 781
  24. 2021Vector Quantized Diffusion Model for Text-to-Image Synthesiscited by 774
  25. 2024Show-o: One Single Transformer to Unify Multimodal Understanding and Generationcited by 752
  26. 2023RAFT: Reward rAnked FineTuning for Generative Foundation Model Alignmentcited by 686
  27. 2023DataComp: In search of the next generation of multimodal datasetscited by 644
  28. 2023NExT-GPT: Any-to-Any Multimodal LLMcited by 615
  29. 2024Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Modelcited by 483
  30. 2023OCRBench: On the Hidden Mystery of OCR in Large Multimodal Modelscited by 476
  31. 2024ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignmentcited by 475
  32. 2024Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generationcited by 457
  33. 2023DPOK: Reinforcement Learning for Fine-tuning Text-to-Image Diffusion Modelscited by 450
  34. 2023Directly Fine-Tuning Diffusion Models on Differentiable Rewardscited by 448
  35. 2023Generative Multimodal Models are In-Context Learnerscited by 446
  36. 2023Aligning Text-to-Image Models using Human Feedbackcited by 422
  37. 2024OmniGen: Unified Image Generationcited by 394
  38. 2025BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Datasetcited by 376

50 most cited of 78 All 78 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 28; 2026: 84.

112 papers.

Sort:

By area:

  1. 2025Qwen-Image Technical Reportcited by 1,044
  2. 2025Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformercited by 277
  3. 2025Emu3.5: Native Multimodal Models are World Learnerscited by 132
  4. 2025Reconstruction Alignment Improves Unified Multimodal Modelscited by 49
  5. 2026Beyond Language Modeling: An Exploration of Multimodal Pretrainingcited by 37
  6. 2026LongCat-Next: Lexicalizing Modalities as Discrete Tokenscited by 36
  7. 2025TUNA: Taming Unified Visual Representations for Native Unified Multimodal Modelscited by 34
  8. 2026HiDream-O1-Image: A Natively Unified Image Generative Foundation Model with Pixel-level Unified Transformercited by 32
  9. 2026SenseNova-U1: Unifying Multimodal Understanding and Generation with NEO-unify Architecturecited by 32
  10. 2026InternVL-U: Democratizing Unified Multimodal Models for Understanding, Reasoning, Generation and Editingcited by 28
  11. 2025UniGenBench++: A Unified Semantic Evaluation Benchmark for Text-to-Image Generationcited by 25
  12. 2026GenAgent: Scaling Text-to-Image Generation via Agentic Multimodal Reasoningcited by 24
38 more
  1. 2026GEMS: Agent-Native Multimodal Generation with Memory and Skillscited by 21
  2. 2025Uni-X: Mitigating Modality Conflict with a Two-End-Separated Architecture for Unified Multimodal Modelscited by 18
  3. 2026Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generationcited by 18
  4. 2025MANZANO: A Simple and Scalable Unified Multimodal Model with a Hybrid Vision Tokenizercited by 15
  5. 2026TextPecker: Rewarding Structural Anomaly Quantification for Enhancing Visual Text Renderingcited by 15
  6. 2026DeepGen 1.0: A Lightweight Unified Multimodal Model for Advancing Image Generation and Editingcited by 14
  7. 2025Unified Multimodal Models as Auto-Encoderscited by 13
  8. 2026Making Reconstruction FID Predictive of Diffusion Generation FIDcited by 13
  9. 2026LLaDA2.0-Uni: Unifying Multimodal Understanding and Generation with Diffusion Large Language Modelcited by 13
  10. 2025Consolidating Reinforcement Learning for Multimodal Discrete Diffusion Modelscited by 12
  11. 2026JoyAI-Image: Awaking Spatial Intelligence in Unified Multimodal Understanding and Generationcited by 12
  12. 2025WiseEdit: Benchmarking Cognition- and Creativity-Informed Image Editingcited by 11
  13. 2026BizGenEval: A Systematic Benchmark for Commercial Visual Content Generationcited by 11
  14. 2026GenClaw: Code-Driven Agentic Image Generationcited by 11
  15. 2025Does Understanding Inform Generation in Unified Multimodal Models? From Analysis to Path Forwardcited by 10
  16. 2025VQRAE: Representation Quantization Autoencoders for Multimodal Understanding, Generation and Reconstructioncited by 10
  17. 2025Reinforcement Learning for Large Model: A Surveycited by 9
  18. 2025UniGen-1.5: Enhancing Image Generation and Editing through Reward Unification in Reinforcement Learningcited by 8
  19. 2025Ovis-Image Technical Reportcited by 8
  20. 2025TextEditBench: Evaluating Reasoning-aware Text Editing Beyond Renderingcited by 8
  21. 2026Elastic Diffusion Transformercited by 8
  22. 2026LatentUM: Unleashing the Potential of Interleaved Cross-Modal Reasoning via a Latent-Space Unified Modelcited by 8
  23. 2026Context Unrolling in Omni Modelscited by 8
  24. 2025Pluggable Pruning with Contiguous Layer Distillation for Diffusion Transformerscited by 7
  25. 2025RubricRL: Simple Generalizable Rewards for Text-to-Image Generationcited by 7
  26. 2026Covo-Audio Technical Reportcited by 7
  27. 2026LaViDa-R1: Advancing Reasoning for Unified Multimodal Diffusion Language Modelscited by 7
  28. 2026Ask, Solve, Generate: Self-Evolving Unified Multimodal Understanding and Generation via Self-Consistency Rewardscited by 7
  29. 2026SenseNova-U1.5: Towards Native Unified Visual Intelligencecited by 7
  30. 2025STAR: STacked AutoRegressive Scheme for Unified Multimodal Learningcited by 6
  31. 2025SemanticGen: Video Generation in Semantic Spacecited by 6
  32. 2026Learning to Generate via Understanding: Understanding-Driven Intrinsic Rewarding for Unified Multimodal Modelscited by 6
  33. 2026CoCo: Code as CoT for Text-to-Image Preview and Rare Concept Generationcited by 6
  34. 2026GlyphPrinter: Region-Grouped Direct Preference Optimization for Glyph-Accurate Visual Text Renderingcited by 6
  35. 2026Lens: Rethinking Training Efficiency for Foundational Text-to-Image Modelscited by 6
  36. 2026ERNIE-Image Technical Reportcited by 6
  37. 2026Boogu-Image-0.1: Boosting Open Agentic Multimodal Generation via Understanding under a Minimal Budgetcited by 6
  38. 2026An Empirical Study of Training Pixel-Space Text-to-Image Diffusion Modelscited by 6

50 most cited of 112 All 112 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025AR-GRPO: Training Autoregressive Image Generation Models via Reinforcement Learning
  2. 2025Selftok: Discrete Visual Tokens of Autoregression, by Diffusion, and for Reasoning
  3. 2025STAGE: Stable and Generalizable GRPO for Autoregressive Image Generation
  4. 2025Beyond Next-Token: Next-X Prediction for Autoregressive Visual Generation
  5. 2024Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation
  6. 2026ARM: An AutoRegressive Large Multimodal Model with Unified Discrete Representations
  7. 2025Bridging Continuous and Discrete Tokens for Autoregressive Visual Generation
  8. 2025Rethinking Discrete Tokens: Treating Them as Conditions for Continuous Autoregressive Image Synthesis