mlsanityData to 14 Oct 2026

Unified Concept Editing in Diffusion Models

arXiv 2308.14761 ↗First posted 25 Aug 2023cs.CV cs.LGcited by 388

On the map: Diffusion models → Robustness, privacy and adversarial aspects

Abstract (arXiv)

Text-to-image models suffer from various safety issues that may limit their suitability for deployment. Previous methods have separately addressed individual issues of bias, copyright, and offensive content in text-to-image models. However, in the real world, all of these issues appear simultaneously in the same model. We present a method that tackles all issues with a single approach. Our method, Unified Concept Editing (UCE), edits the model without training using a closed-form solution, and scales seamlessly to concurrent edits on text-conditional diffusion models. We demonstrate scalable simultaneous debiasing, style erasure, and content moderation by editing text-to-image projections, and we present extensive experiments demonstrating improved efficacy and scalability over prior work. Our code is available at https://unified.baulab.info

Key facts. Cited by 388 papers in the map (data through 14 Oct 2026). Placed in Robustness, privacy and adversarial aspects, one of 14 directions in Diffusion models. Builds on 34 of its 52 references that are in the map. First posted on arXiv 25 Aug 2023.

Builds on

34 of 52 references.

Closest ancestors · 8

Same direction, newest first

  1. 2023Towards Safe Self-Distillation of Internet-Scale Text-to-Image Diffusion Models12 Jul
  2. 2023Forget-Me-Not: Learning to Forget in Text-to-Image Diffusion Models30 Mar
  3. 2023Ablating Concepts in Text-to-Image Diffusion Models23 Mar
  4. 2023Erasing Concepts from Diffusion Models13 Mar
  5. 2023Raising the Cost of Malicious AI-Powered Image Editing13 Feb
  6. 2023Extracting Training Data from Diffusion Models30 Jan
  7. 2022Safe Latent Diffusion: Mitigating Inappropriate Degeneration in Diffusion Models9 Nov
  8. 2022Red-Teaming the Stable Diffusion Safety Filter3 Oct

Other references · 26

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  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. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  7. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  8. 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Modelscited by 3,614
  9. 2022Locating and Editing Factual Associations in GPTcited by 2,691
  10. 2022Mass-Editing Memory in a Transformercited by 1,001
  11. 2018Gender Bias in Coreference Resolution: Evaluation and Debiasing Methodscited by 659
  12. 2022Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Modelscited by 424
14 more
  1. 2022Easily Accessible Text-to-Image Generation Amplifies Demographic Stereotypes at Large Scalecited by 262
  2. 2022DALL-Eval: Probing the Reasoning Skills and Social Biases of Text-to-Image Generation Modelscited by 230
  3. 2023Stable Bias: Analyzing Societal Representations in Diffusion Modelscited by 226
  4. 2023Understanding and Mitigating Copying in Diffusion Modelscited by 209
  5. 2023Glaze: Protecting Artists from Style Mimicry by Text-to-Image Modelscited by 205
  6. 2020fastai: A Layered API for Deep Learningcited by 183
  7. 2023Selective Amnesia: A Continual Learning Approach to Forgetting in Deep Generative Modelscited by 177
  8. 2023Debiasing Vision-Language Models via Biased Promptscited by 149
  9. 2022The Privacy Onion Effect: Memorization is Relativecited by 136
  10. 2023Editing Implicit Assumptions in Text-to-Image Diffusion Modelscited by 130
  11. 2020Rewriting a Deep Generative Modelcited by 107
  12. 2023The Bias Amplification Paradox in Text-to-Image Generationcited by 70
  13. 2023ReFACT: Updating Text-to-Image Models by Editing the Text Encodercited by 32
  14. 2023A Friendly Face: Do Text-to-Image Systems Rely on Stereotypes when the Input is Under-Specified?cited by 25

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 14; 2024: 94; 2025: 138; 2026: 142.

388 papers.

Sort:

By area:

  1. 2024MACE: Mass Concept Erasure in Diffusion Modelscited by 307
  2. 2023To Generate or Not? Safety-Driven Unlearned Diffusion Models Are Still Easy To Generate Unsafe Images ... For Nowcited by 210
  3. 2024Defensive Unlearning with Adversarial Training for Robust Concept Erasure in Diffusion Modelscited by 168
  4. 2023Concept Sliders: LoRA Adaptors for Precise Control in Diffusion Modelscited by 153
  5. 2024Reliable and Efficient Concept Erasure of Text-to-Image Diffusion Modelscited by 133
  6. 2024SAFREE: Training-Free and Adaptive Guard for Safe Text-to-Image And Video Generationcited by 133
  7. 2023Receler: Reliable Concept Erasing of Text-to-Image Diffusion Models via Lightweight Eraserscited by 120
  8. 2024EraseAnything: Enabling Concept Erasure in Rectified Flow Transformerscited by 95
  9. 2023Finetuning Text-to-Image Diffusion Models for Fairnesscited by 92
  10. 2025SAeUron: Interpretable Concept Unlearning in Diffusion Models with Sparse Autoencoderscited by 80
  11. 2024Jailbreaking Prompt Attack: A Controllable Adversarial Attack against Diffusion Modelscited by 72
  12. 2023Self-Discovering Interpretable Diffusion Latent Directions for Responsible Text-to-Image Generationcited by 69
38 more
  1. 2024Unlearning Concepts in Diffusion Model via Concept Domain Correction and Concept Preserving Gradientcited by 69
  2. 2023Toward effective protection against diffusion based mimicry through score distillationcited by 65
  3. 2025Set You Straight: Auto-Steering Denoising Trajectories to Sidestep Unwanted Conceptscited by 62
  4. 2024OpenBias: Open-set Bias Detection in Text-to-Image Generative Modelscited by 61
  5. 2024Erasing Undesirable Influence in Diffusion Modelscited by 57
  6. 2024ConceptPrune: Concept Editing in Diffusion Models via Skilled Neuron Pruningcited by 55
  7. 2024Scissorhands: Scrub Data Influence via Connection Sensitivity in Networkscited by 52
  8. 2024UnlearnCanvas: Stylized Image Dataset for Enhanced Machine Unlearning Evaluation in Diffusion Modelscited by 50
  9. 2024Direct Unlearning Optimization for Robust and Safe Text-to-Image Modelscited by 50
  10. 2024Erasing Undesirable Concepts in Diffusion Models with Adversarial Preservationcited by 50
  11. 2024Machine Unlearning in Generative AI: A Surveycited by 49
  12. 2025T2ISafety: Benchmark for Assessing Fairness, Toxicity, and Privacy in Image Generationcited by 48
  13. 2024Challenging Forgets: Unveiling the Worst-Case Forget Sets in Machine Unlearningcited by 44
  14. 2024Dissecting Language Models: Machine Unlearning via Selective Pruningcited by 42
  15. 2024STEREO: A Two-Stage Framework for Adversarially Robust Concept Erasing from Text-to-Image Diffusion Modelscited by 42
  16. 2024Meta-Unlearning on Diffusion Models: Preventing Relearning Unlearned Conceptscited by 40
  17. 2024Precise, Fast, and Low-cost Concept Erasure in Value Space: Orthogonal Complement Matterscited by 40
  18. 2024Finding NeMo: Localizing Neurons Responsible For Memorization in Diffusion Modelscited by 37
  19. 2025Fantastic Targets for Concept Erasure in Diffusion Models and Where To Find Themcited by 37
  20. 2025SPEED: Scalable, Precise, and Efficient Concept Erasure for Diffusion Modelscited by 36
  21. 2024Adversarial Attacks and Defenses on Text-to-Image Diffusion Models: A Surveycited by 35
  22. 2024FairRAG: Fair Human Generation via Fair Retrieval Augmentationcited by 34
  23. 2024Erasing Conceptual Knowledge from Language Modelscited by 34
  24. 2025Localized Concept Erasure for Text-to-Image Diffusion Models Using Training-Free Gated Low-Rank Adaptationcited by 33
  25. 2024Jailbreak Attacks and Defenses against Multimodal Generative Models: A Surveycited by 32
  26. 2023Fair Text-to-Image Diffusion via Fair Mappingcited by 31
  27. 2024Erasing Concepts from Text-to-Image Diffusion Models with Few-shot Unlearningcited by 29
  28. 2024T2IShield: Defending Against Backdoors on Text-to-Image Diffusion Modelscited by 29
  29. 2024Dynamic Negative Guidance of Diffusion Modelscited by 28
  30. 2024Robust Concept Erasure Using Task Vectorscited by 27
  31. 2024Fuzz-Testing Meets LLM-Based Agents: An Automated and Efficient Framework for Jailbreaking Text-To-Image Generation Modelscited by 27
  32. 2025CURE: Concept Unlearning via Orthogonal Representation Editing in Diffusion Modelscited by 27
  33. 2024TraSCE: Trajectory Steering for Concept Erasurecited by 26
  34. 2023Stable Diffusion Exposed: Gender Bias from Prompt to Imagecited by 25
  35. 2025Fine-Grained Erasure in Text-to-Image Diffusion-based Foundation Modelscited by 25
  36. 2024The Quest for the Right Mediator: Surveying Mechanistic Interpretability Through the Lens of Causal Mediation Analysiscited by 24
  37. 2024Pruning for Robust Concept Erasing in Diffusion Modelscited by 22
  38. 2024FAIntbench: A Holistic and Precise Benchmark for Bias Evaluation in Text-to-Image Modelscited by 22

50 most cited of 388 All 388 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026Erasing Thousands of Concepts: Towards Scalable and Practical Concept Erasure for Text-to-Image Diffusion Models
  2. 2026Compositional Concept Erasure in Text-to-Image Diffusion Models via Hierarchically Grounded Semantic Surgery
  3. 2023All but One: Surgical Concept Erasing with Model Preservation in Text-to-Image Diffusion Models
  4. 2025ACE: Anti-Editing Concept Erasure in Text-to-Image Models
  5. 2023Custom-Edit: Text-Guided Image Editing with Customized Diffusion Models
  6. 2023Towards Real-time Text-driven Image Manipulation with Unconditional Diffusion Models
  7. 2023PRedItOR: Text Guided Image Editing with Diffusion Prior
  8. 2026i1: A Simple and Fully Open Recipe for Strong Text-to-Image Models