mlsanityData to 14 Oct 2026

ACE: Concept Editing in Diffusion Models without Performance Degradation

arXiv 2503.08116 ↗First posted 11 Mar 2025cs.CVcited by 6

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

Abstract (arXiv)

Diffusion-based text-to-image models have demonstrated remarkable capabilities in generating realistic images, but they raise societal and ethical concerns, such as the creation of unsafe content. While concept editing is proposed to address these issues, they often struggle to balance the removal of unsafe concept with maintaining the model's general genera-tive capabilities. In this work, we propose ACE, a new editing method that enhances concept editing in diffusion models. ACE introduces a novel cross null-space projection approach to precisely erase unsafe concept while maintaining the model's ability to generate high-quality, semantically consistent images. Extensive experiments demonstrate that ACE significantly outperforms the advancing baselines,improving semantic consistency by 24.56% and image generation quality by 34.82% on average with only 1% of the time cost. These results highlight the practical utility of concept editing by mitigating its potential risks, paving the way for broader applications in the field. Code is avaliable at https://github.com/littlelittlenine/ACE-zero.git

Key facts. Cited by 6 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 29 of its 46 references that are in the map. First posted on arXiv 11 Mar 2025.

Builds on

29 of 46 references.

Closest ancestors · 9

Same direction, newest first

  1. 2024ConceptPrune: Concept Editing in Diffusion Models via Skilled Neuron Pruning29 May
  2. 2024Pruning for Robust Concept Erasing in Diffusion Models26 May
  3. 2024Defensive Unlearning with Adversarial Training for Robust Concept Erasure in Diffusion Models24 May
  4. 2023Receler: Reliable Concept Erasing of Text-to-Image Diffusion Models via Lightweight Erasers29 Nov
  5. 2023SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation19 Oct
  6. 2023Unified Concept Editing in Diffusion Models25 Aug
  7. 2023Extracting Training Data from Diffusion Models30 Jan
  8. 2022Safe Latent Diffusion: Mitigating Inappropriate Degeneration in Diffusion Models9 Nov
  9. 2022Red-Teaming the Stable Diffusion Safety Filter3 Oct

Other references · 20

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2020Language Models are Few-Shot Learnerscited by 32,419
  4. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  5. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  6. 2020Denoising Diffusion Implicit Modelscited by 9,781
  7. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  8. 2022Locating and Editing Factual Associations in GPTcited by 2,691
  9. 2022Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Modelscited by 424
  10. 2024AlphaEdit: Null-Space Constrained Knowledge Editing for Language Modelscited by 236
  11. 2022DALL-Eval: Probing the Reasoning Skills and Social Biases of Text-to-Image Generation Modelscited by 230
  12. 2023Debiasing Vision-Language Models via Biased Promptscited by 149
8 more
  1. 2021Training Networks in Null Space of Feature Covariance for Continual Learningcited by 143
  2. 2025AnyEdit: Edit Any Knowledge Encoded in Language Modelscited by 76
  3. 2025Reinforced Lifelong Editing for Language Modelscited by 34
  4. 2022Exploiting Cultural Biases via Homoglyphs in Text-to-Image Synthesiscited by 29
  5. 2024Robust Concept Erasure Using Task Vectorscited by 27
  6. 2024Enhancing Zero-Shot Vision Models by Label-Free Prompt Distribution Learning and Bias Correctingcited by 26
  7. 2024Selective Vision-Language Subspace Projection for Few-shot CLIPcited by 18
  8. 2024Boosting Few-Shot Learning via Attentive Feature Regularizationcited by 10

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 4.

6 papers.

  1. 2025Set You Straight: Auto-Steering Denoising Trajectories to Sidestep Unwanted Conceptscited by 62
  2. 2026Are We Evaluating the Edit Locality of LLM Model Editing Properly?cited by 4
  3. 2025ACE: Attentional Concept Erasure in Diffusion Modelscited by 3
  4. 2026Principled Steering via Null-space Projection for Jailbreak Defense in Vision-Language Modelscited by 2
  5. 2026From Backward Spreading to Forward Replay: Revisiting Target Construction in LLM Parameter Editingcited by 2
  6. 2026Concept Unlearning via Cross-Attention Activation Projection for Diffusion Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025ACE: Anti-Editing Concept Erasure in Text-to-Image Models
  2. 2024Editing Massive Concepts in Text-to-Image Diffusion Models
  3. 2025SPEED: Scalable, Precise, and Efficient Concept Erasure for Diffusion Models
  4. 2025Semantic Surgery: Zero-Shot Concept Erasure in Diffusion Models
  5. 2022DiffEdit: Diffusion-based semantic image editing with mask guidance
  6. 2024ACE: All-round Creator and Editor Following Instructions via Diffusion Transformer
  7. 2023All but One: Surgical Concept Erasing with Model Preservation in Text-to-Image Diffusion Models
  8. 2023Custom-Edit: Text-Guided Image Editing with Customized Diffusion Models