mlsanityData to 14 Oct 2026

Edit Flows: Flow Matching with Edit Operations

arXiv 2506.09018 ↗First posted 10 Jun 2025cs.LG cs.AIcited by 55

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

Abstract (arXiv)

Autoregressive generative models naturally generate variable-length sequences, while non-autoregressive models struggle, often imposing rigid, token-wise structures. We propose Edit Flows, a non-autoregressive model that overcomes these limitations by defining a discrete flow over sequences through edit operationsunicodex2013insertions, deletions, and substitutions. By modeling these operations within a Continuous-time Markov Chain over the sequence space, Edit Flows enable flexible, position-relative generation that aligns more closely with the structure of sequence data. Our training method leverages an expanded state space with auxiliary variables, making the learning process efficient and tractable. Empirical results show that Edit Flows outperforms both autoregressive and mask models on image captioning and significantly outperforms the mask construction in text and code generation.

Key facts. Cited by 55 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 51 of its 57 references that are in the map. First posted on arXiv 10 Jun 2025.

Builds on

51 of 57 references.

References · 51

  1. 2017Attention Is All You Needcited by 60,361
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  3. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  4. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  5. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  6. 2024The Llama 3 Herd of Modelscited by 14,715
  7. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 13,028
  8. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  9. 2022Scalable Diffusion Models with Transformerscited by 6,054
  10. 2022Classifier-Free Diffusion Guidancecited by 5,847
  11. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,703
  12. 2023Improved Baselines with Visual Instruction Tuningcited by 4,418
38 more
  1. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  2. 2019XLNet: Generalized Autoregressive Pretraining for Language Understandingcited by 2,811
  3. 2023Code Llama: Open Foundation Models for Codecited by 2,139
  4. 2021Structured Denoising Diffusion Models in Discrete State-Spacescited by 1,719
  5. 2021Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolationcited by 1,103
  6. 2022MaskGIT: Masked Generative Image Transformercited by 1,099
  7. 2025Large Language Diffusion Modelscited by 912
  8. 2024Simple and Effective Masked Diffusion Language Modelscited by 848
  9. 2023Discrete Diffusion Modeling by Estimating the Ratios of the Data Distributioncited by 745
  10. 2024Simplified and Generalized Masked Diffusion for Discrete Datacited by 626
  11. 2021ClipCap: CLIP Prefix for Image Captioningcited by 559
  12. 2019Unified Vision-Language Pre-Training for Image Captioning and VQAcited by 494
  13. 2022A Continuous Time Framework for Discrete Denoising Modelscited by 451
  14. 2024Your Absorbing Discrete Diffusion Secretly Models the Conditional Distributions of Clean Datacited by 381
  15. 2024Discrete Flow Matchingcited by 317
  16. 2024LongRoPE: Extending LLM Context Window Beyond 2 Million Tokenscited by 313
  17. 2024Generative Flows on Discrete State-Spaces: Enabling Multimodal Flows with Applications to Protein Co-Designcited by 305
  18. 2024Flow Matching Guide and Codecited by 301
  19. 2022Generating Sequences by Learning to Self-Correctcited by 277
  20. 2021Autoregressive Diffusion Modelscited by 210
  21. 2024Masked Diffusion Models are Secretly Time-Agnostic Masked Models and Exploit Inaccurate Categorical Samplingcited by 208
  22. 2019Levenshtein Transformercited by 199
  23. 2018Deterministic Non-Autoregressive Neural Sequence Modeling by Iterative Refinementcited by 195
  24. 2024Flex Attention: A Programming Model for Generating Optimized Attention Kernelscited by 185
  25. 2022COLD Decoding: Energy-based Constrained Text Generation with Langevin Dynamicscited by 161
  26. 2022CM3: A Causal Masked Multimodal Model of the Internetcited by 142
  27. 2024Unlocking Guidance for Discrete State-Space Diffusion and Flow Modelscited by 129
  28. 2019Insertion Transformer: Flexible Sequence Generation via Insertion Operationscited by 125
  29. 2018CGMH: Constrained Sentence Generation by Metropolis-Hastings Samplingcited by 94
  30. 2024Generator Matching: Generative modeling with arbitrary Markov processescited by 86
  31. 2020Imputer: Sequence Modelling via Imputation and Dynamic Programmingcited by 66
  32. 2024Flow Matching with General Discrete Paths: A Kinetic-Optimal Perspectivecited by 57
  33. 2019Insertion-based Decoding with automatically Inferred Generation Ordercited by 43
  34. 2023On Kinetic Optimal Probability Paths for Generative Modelscited by 42
  35. 2022DiffusER: Discrete Diffusion via Edit-based Reconstructioncited by 41
  36. 2018Lecture Notes: Temporal Point Processes and the Conditional Intensity Functioncited by 40
  37. 2021Relative Positional Encoding for Transformers with Linear Complexitycited by 37
  38. 2023Trans-Dimensional Generative Modeling via Jump Diffusion Modelscited by 35

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 21; 2026: 34.

55 papers.

Sort:

By area:

  1. 2025Any-Order Flexible Length Masked Diffusioncited by 51
  2. 2025Don't Settle Too Early: Self-Reflective Remasking for Diffusion Language Modelscited by 37
  3. 2025Continuously Augmented Discrete Diffusion model for Categorical Generative Modelingcited by 34
  4. 2026DreamOn: Diffusion Language Models For Code Infilling Beyond Fixed-size Canvascited by 19
  5. 2025Dream-VL & Dream-VLA: Open Vision-Language and Vision-Language-Action Models with Diffusion Language Model Backbonecited by 18
  6. 2025OneFlow: Concurrent Mixed-Modal and Interleaved Generation with Edit Flowscited by 17
  7. 2025Unveiling the Potential of Diffusion Large Language Model in Controllable Generationcited by 15
  8. 2025Planner Aware Path Learning in Diffusion Language Models Trainingcited by 14
  9. 2026Discrete Adjoint Matchingcited by 11
  10. 2026Generation Order and Parallel Decoding in Masked Diffusion Models: An Information-Theoretic Perspectivecited by 10
  11. 2025Masks Can Be Distracting: On Context Comprehension in Diffusion Language Modelscited by 9
  12. 2025Edit-Based Flow Matching for Temporal Point Processescited by 8
38 more
  1. 2025Exploring Synthesizable Chemical Space with Iterative Pathway Refinementscited by 7
  2. 2025Discrete Diffusion for Generative Modeling of Text-Aligned Speech Tokenscited by 7
  3. 2025A Diffusion Model to Shrink Proteins While Maintaining Their Functioncited by 7
  4. 2026Beyond Masks: Efficient, Flexible Diffusion Language Models via Deletion-Insertion Processescited by 7
  5. 2025Discrete Guidance Matching: Exact Guidance for Discrete Flow Matchingcited by 6
  6. 2025Guided Star-Shaped Masked Diffusioncited by 5
  7. 2025Branching Flows: Discrete, Continuous, and Manifold Flow Matching with Splits and Deletionscited by 5
  8. 2026DFM-VLA: Iterative Action Refinement for Robot Manipulation via Discrete Flow Matchingcited by 5
  9. 2025Rethinking Reasoning with MDLMs: Early Exits, Post-hoc Reasoning, and Beyondcited by 4
  10. 2025XLM: A Python package for non-autoregressive language modelscited by 4
  11. 2025Drax: Speech Recognition with Discrete Flow Matchingcited by 3
  12. 2026Towards A Generative Protein Evolution Machine with DPLM-Evocited by 3
  13. 2025DiffuTester: Accelerating Unit Test Generation for Diffusion LLMs via Mining Structural Patterncited by 2
  14. 2025Time dependent loss reweighting for flow matching and diffusion models is theoretically justifiedcited by 2
  15. 2025Foundations of Diffusion Models in General State Spaces: A Self-Contained Introductioncited by 2
  16. 2026Diffusion LMs Can Approximate Optimal Infilling Lengths Implicitlycited by 2
  17. 2026Edit-Based Refinement for Parallel Masked Diffusion Language Modelscited by 2
  18. 2026Generative Molecular Morphing for Flexible-Size Design via Unbalanced Optimal Transportcited by 2
  19. 2026A2D2: Fine-Tuning Any-Length Discrete Diffusion for Adaptive Decodingcited by 2
  20. 2026Distilled Continuous Diffusion Language Models Can Write Code in Few Steps---or Onecited by 2
  21. 2026Insertion Based Sequence Generation with Learnable Order Dynamicscited by 1
  22. 2026Variable-Length Generative Protein Design via Generalized Poisson Flowcited by 1
  23. 2026Multi-Mask Diffusion Language Models for Few-Step Generationcited by 1
  24. 2026Expanding Flow Mapscited by 1
  25. 2026From Interface to Inference: Eliciting Any-Order Inference from Any-Order Modelscited by 1
  26. 2026Mechanistic Reaction Prediction via Discrete Flow Matching on Graph-Structured Electron Occupationcited by 1
  27. 2026ELF-REG: Scaling Continuous Diffusion Language Models to Reasoning Taskscited by 1
  28. 2026Reasoning with Continuous Latent Diffusioncited by 1
  29. 2026Reconsidering Positional Supervision in Masked Diffusion Language Model Trainingcited by 0
  30. 2026Tree-Conditioned Edit Flows for Ancestral Sequence Reconstructioncited by 0
  31. 2026LPDP: Inference-Time Reward Control for Variable-Length DNA Generation with Edit Flowscited by 0
  32. 2026Latent Process Generator Matchingcited by 0
  33. 2026EditSR: Enhancing Neural Symbolic Regression via Edit-based Rectificationcited by 0
  34. 2026A Continuous-Time Markov Chain Framework for Insertion Language Modelscited by 0
  35. 2026Flexible Flows for Biological Sequence Designcited by 0
  36. 2026Nemotron-Labs-Diffusion-Image: Advancing Masked Discrete Diffusion for High-Resolution Image Synthesiscited by 0
  37. 2026CARVE: Verified Expansion for Variable-Length Generation in Diffusion Language Modelscited by 0
  38. 2026HorizonFlow: Variable-Length Planning for Offline Goal-Conditioned RLcited by 0

50 most cited of 55 All 55 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026Bootstrap Your Generator: Unpaired Visual Editing with Flow Matching
  2. 2025SplitFlow: Flow Decomposition for Inversion-Free Text-to-Image Editing
  3. 2026Shifting the Breaking Point of Flow Matching for Multi-Instance Editing
  4. 2026DiTailed: Ensuring Visual Object Consistency in Text-Image-to-Image Flow Matching Models
  5. 2025EditAR: Unified Conditional Generation with Autoregressive Models
  6. 2025NextStep-1: Toward Autoregressive Image Generation with Continuous Tokens at Scale
  7. 2026Constrained Edit Fields for Training-Free Flow Editing
  8. 2023Latent Space Editing in Transformer-Based Flow Matching