mlsanityData to 14 Oct 2026

Modality-Decoupled Online Recursive Editing

arXiv 2605.20273 ↗First posted 19 May 2026cs.LG cs.AIcited by 1

On the map: Multimodal LLMs / vision-language models → Image editing and quality assessment

Abstract (arXiv)

Online model editing for multimodal large language models (MLLMs) requires assimilating a stream of corrections under tight compute and memory budgets. Yet editors developed for text-only LLMs often degrade on MLLMs: visually dominant activations skew the statistics that shape updates, causing cross-modal conflict, while sequential writes become entangled in a shared edit space and amplify long-horizon interference, causing inter-edit interference. To address these, we propose M-ORE, a modality-decoupled online recursive editor for lifelong MLLM adaptation. M-ORE is derived from a unified proximal-projection formulation and admits a closed-form update with a Sherman-Morrison recursion, yielding constant per-edit overhead. It maintains module-wise locality statistics for the text stack and the visual projector to avoid visually dominated update shaping and performs continual updates in a fixed orthogonal low-rank edit subspace via a Sherman-Morrison recursion to mitigate long-horizon interference. Experiments on multiple MLLM backbones and online editing benchmarks show that our M-ORE method consistently improves reliability, generality, and locality over strong baselines, while achieving favorable quality-efficiency scaling. Our code is publicly available at https://github.com/lab-klc/M-ORE.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Image editing and quality assessment, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 46 of its 51 references that are in the map. First posted on arXiv 19 May 2026.

Builds on

46 of 51 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025DualEdit: Dual Editing for Knowledge Updating in Vision-Language Models16 Jun
  2. 2024Lifelong Knowledge Editing for Vision Language Models with Low-Rank Mixture-of-Experts23 Nov
  3. 2024Attribution Analysis Meets Model Editing: Advancing Knowledge Correction in Vision Language Models with VisEdit19 Aug

Other references · 43

  1. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  2. 2023GPT-4 Technical Reportcited by 18,396
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  4. 2023Improved Baselines with Visual Instruction Tuningcited by 4,418
  5. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  6. 2022OPT: Open Pre-trained Transformer Language Modelscited by 3,512
  7. 2025Qwen3-VL Technical Reportcited by 2,793
  8. 2022Locating and Editing Factual Associations in GPTcited by 2,691
  9. 2023A Survey of Large Language Modelscited by 2,576
  10. 2016Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answeringcited by 2,516
  11. 2023MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Modelscited by 1,652
  12. 2015Microsoft COCO Captions: Data Collection and Evaluation Servercited by 1,462
31 more
  1. 2023Video-LLaVA: Learning United Visual Representation by Alignment Before Projectioncited by 1,402
  2. 2020Transformer Feed-Forward Layers Are Key-Value Memoriescited by 1,336
  3. 2019OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledgecited by 1,121
  4. 2021GLM: General Language Model Pretraining with Autoregressive Blank Infillingcited by 1,031
  5. 2022Mass-Editing Memory in a Transformercited by 1,001
  6. 2021Knowledge Neurons in Pretrained Transformerscited by 601
  7. 2021Editing Factual Knowledge in Language Modelscited by 553
  8. 2021Fast Model Editing at Scalecited by 498
  9. 2022Memory-Based Model Editing at Scalecited by 447
  10. 2023Editing Large Language Models: Problems, Methods, and Opportunitiescited by 372
  11. 2023Can We Edit Factual Knowledge by In-Context Learning?cited by 285
  12. 2022Aging with GRACE: Lifelong Model Editing with Discrete Key-Value Adaptorscited by 271
  13. 2024AlphaEdit: Null-Space Constrained Knowledge Editing for Language Modelscited by 236
  14. 2023Transformer-Patcher: One Mistake worth One Neuroncited by 197
  15. 2024A Comprehensive Study of Knowledge Editing for Large Language Modelscited by 163
  16. 2020Editable Neural Networkscited by 155
  17. 2024Model Editing Harms General Abilities of Large Language Models: Regularization to the Rescuecited by 103
  18. 2023MELO: Enhancing Model Editing with Neuron-Indexed Dynamic LoRAcited by 80
  19. 2023Can We Edit Multimodal Large Language Models?cited by 44
  20. 2024Lifelong Knowledge Editing for LLMs with Retrieval-Augmented Continuous Prompt Learningcited by 40
  21. 2024LLMs Meet Multimodal Generation and Editing: A Surveycited by 35
  22. 2025Reinforced Lifelong Editing for Language Modelscited by 34
  23. 2025MedEyes: Learning Dynamic Visual Focus for Medical Progressive Diagnosiscited by 32
  24. 2025WikiBigEdit: Understanding the Limits of Lifelong Knowledge Editing in LLMscited by 26
  25. 2024Neuron-Level Sequential Editing for Large Language Modelscited by 20
  26. 2024Everything is Editable: Extend Knowledge Editing to Unstructured Data in Large Language Modelscited by 19
  27. 2025Robust Multimodal Large Language Models Against Modality Conflictcited by 18
  28. 2026ChordEdit: One-Step Low-Energy Transport for Image Editingcited by 18
  29. 2026Render-in-the-Loop: Vector Graphics Generation via Visual Self-Feedbackcited by 12
  30. 2026MMErroR: A Benchmark for Erroneous Reasoning in Vision-Language Modelscited by 10
  31. 2025On the Superimposed Noise Accumulation Problem in Sequential Knowledge Editing of Large Language Modelscited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Multimodal Knowledge Edit-Scoped Generalization for Online Recursive MLLM Editingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Generalizable Multimodal Large Language Model Editing via Invariant Trajectory Learning
  2. 2025Rethinking Residual Distribution in Locate-then-Edit Model Editing
  3. 2024LEMoE: Advanced Mixture of Experts Adaptor for Lifelong Model Editing of Large Language Models
  4. 2026Evaluating and Understanding Model Editing for Medical Vision Language Models
  5. 2026Beyond the Covariance Trap: Unlocking Generalization in Same-Subject Knowledge Editing for Large Language Models
  6. 2025InComeS: Integrating Compression and Selection Mechanisms into LLMs for Efficient Model Editing
  7. 2026CrispEdit: Low-Curvature Projections for Scalable Non-Destructive LLM Editing
  8. 2026Correct When Paired, Wrong When Split: Decoupling and Editing Modality-Specific Neurons in MLLMs