mlsanityData to 14 Oct 2026

Fantastic Targets for Concept Erasure in Diffusion Models and Where To Find Them

arXiv 2501.18950 ↗First posted 31 Jan 2025cs.LG cs.AI cs.CVcited by 37

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

also in 1 more
Abstract (arXiv)

Concept erasure has emerged as a promising technique for mitigating the risk of harmful content generation in diffusion models by selectively unlearning undesirable concepts. The common principle of previous works to remove a specific concept is to map it to a fixed generic concept, such as a neutral concept or just an empty text prompt. In this paper, we demonstrate that this fixed-target strategy is suboptimal, as it fails to account for the impact of erasing one concept on the others. To address this limitation, we model the concept space as a graph and empirically analyze the effects of erasing one concept on the remaining concepts. Our analysis uncovers intriguing geometric properties of the concept space, where the influence of erasing a concept is confined to a local region. Building on this insight, we propose the Adaptive Guided Erasure (AGE) method, which dynamically selects optimal target concepts tailored to each undesirable concept, minimizing unintended side effects. Experimental results show that AGE significantly outperforms state-of-the-art erasure methods on preserving unrelated concepts while maintaining effective erasure performance. Our code is published at https://github.com/tuananhbui89/Adaptive-Guided-Erasure.

Key facts. Cited by 37 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 27 of its 32 references that are in the map. First posted on arXiv 31 Jan 2025.

Builds on

27 of 32 references.

Closest ancestors · 11

Same direction, newest first

  1. 2024Erasing Undesirable Concepts in Diffusion Models with Adversarial Preservation21 Oct
  2. 2024Hiding and Recovering Knowledge in Text-to-Image Diffusion Models via Learnable Prompts18 Mar
  3. 2024MACE: Mass Concept Erasure in Diffusion Models10 Mar
  4. 2023One-Dimensional Adapter to Rule Them All: Concepts, Diffusion Models and Erasing Applications26 Dec
  5. 2023Unified Concept Editing in Diffusion Models25 Aug
  6. 2023SneakyPrompt: Jailbreaking Text-to-image Generative Models20 May
  7. 2023Forget-Me-Not: Learning to Forget in Text-to-Image Diffusion Models30 Mar
  8. 2023Ablating Concepts in Text-to-Image Diffusion Models23 Mar
  9. 2023Erasing Concepts from Diffusion Models13 Mar
  10. 2022Safe Latent Diffusion: Mitigating Inappropriate Degeneration in Diffusion Models9 Nov
  11. 2022Red-Teaming the Stable Diffusion Safety Filter3 Oct

Other references · 16

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  4. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  5. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  6. 2022Hierarchical Text-Conditional Image Generation with CLIP Latentscited by 6,585
  7. 2021Zero-Shot Text-to-Image Generationcited by 4,288
  8. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
  9. 2022An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversioncited by 2,245
  10. 2017Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Modelscited by 462
  11. 2023Editing Implicit Assumptions in Text-to-Image Diffusion Modelscited by 130
  12. 2023Circumventing Concept Erasure Methods For Text-to-Image Generative Modelscited by 85
4 more
  1. 2020Improving Image Captioning with Better Use of Captionscited by 84
  2. 2022A Unified Wasserstein Distributional Robustness Framework for Adversarial Trainingcited by 36
  3. 2020Improving Adversarial Robustness by Enforcing Local and Global Compactnesscited by 11
  4. 2021Understanding and Achieving Efficient Robustness with Adversarial Supervised Contrastive Learningcited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 15; 2026: 22.

37 papers.

Sort:

By area:

  1. 2025Set You Straight: Auto-Steering Denoising Trajectories to Sidestep Unwanted Conceptscited by 62
  2. 2025Erasing Concepts, Steering Generations: A Comprehensive Survey of Concept Suppressioncited by 10
  3. 2025Continual Unlearning for Text-to-Image Diffusion Models: A Regularization Perspectivecited by 10
  4. 2025Erased but Not Forgotten: How Backdoors Compromise Concept Erasurecited by 9
  5. 2025Erased or Dormant? Rethinking Concept Erasure Through Reversibilitycited by 9
  6. 2025EraseFlow: Learning Concept Erasure Policies via GFlowNet-Driven Alignmentcited by 9
  7. 2025Sparse Autoencoder as a Zero-Shot Classifier for Concept Erasing in Text-to-Image Diffusion Modelscited by 8
  8. 2025CRCE: Coreference-Retention Concept Erasure in Text-to-Image Diffusion Modelscited by 6
  9. 2025Locality-Aware Continual Unlearning for Diffusion Modelscited by 5
  10. 2026AEGIS: Adversarial Target-Guided Retention-Data-Free Robust Concept Erasure from Diffusion Modelscited by 5
  11. 2026Co-occurring Associated REtained concepts in Diffusion Unlearningcited by 4
  12. 2026Closed-Form Concept Erasure via Double Projectionscited by 3
25 more
  1. 2026To Erase, or Not to Erase: Robust Training-Free Concept Erasure with Preservation aware Adaptive Ranked Subspace Expansioncited by 3
  2. 2025VCE: Safe Autoregressive Image Generation via Visual Contrast Exploitationcited by 2
  3. 2025A Unified Framework for Diffusion Model Unlearning with f-Divergencecited by 2
  4. 2026Orthogonal Concept Erasure for Diffusion Modelscited by 2
  5. 2025Automating Evaluation of Diffusion Model Unlearning with (Vision-) Language Model World Knowledgecited by 1
  6. 2025Beyond Fixed Anchors: Precisely Erasing Concepts with Sibling Exclusive Counterpartscited by 1
  7. 2025Rethinking Robust Adversarial Concept Erasure in Diffusion Modelscited by 1
  8. 2026The Illusion of Forgetting: Attack Unlearned Diffusion via Initial Latent Variable Optimizationcited by 1
  9. 2026Roots Beneath the Cut: Uncovering the Risk of Concept Revival in Pruning-Based Unlearning for Diffusion Modelscited by 1
  10. 2026Unlearning the Unpromptable: Prompt-free Instance Unlearning in Diffusion Modelscited by 1
  11. 2026Erasure or Erosion? Evaluating Compositional Degradation in Unlearned Text-To-Image Diffusion Modelscited by 1
  12. 2026Empty SPACE: Cross-Attention Sparsity for Concept Erasure in Diffusion Modelscited by 1
  13. 2026FADE: Frame-Aware Diffusion-Transformer-based Multi-Concept Erasure for Video Unlearningcited by 1
  14. 2026No Concept Escapes the Audit: Auditing-Aware Unlearning for Verifiable Concept Erasure in Diffusion Modelscited by 1
  15. 2026Your Unlearning Gives You Away: Identifying Erased Concepts in Diffusion Modelscited by 1
  16. 2025Mitigating Semantic Collapse in Generative Personalization with Test-Time Embedding Adjustmentcited by 0
  17. 2026Consistency-Preserving Concept Erasure via Unsafe-Safe Pairing and Directional Fisher-weighted Adaptationcited by 0
  18. 2026Concept Removal for Frontier Image Generative Modelscited by 0
  19. 2026Efficient bias mitigation in T2I diffusion models using Concept Graphscited by 0
  20. 2026TILDE: TILt-based Distributional Erasure for Concept Unlearningcited by 0
  21. 2026Uni-AdaVD: Universal Concept Erasure for Visual Generation via Orthogonal Value Decompositioncited by 0
  22. 2026MapRoute++: Surrogate-Guided Semantic Routing for Visual Concept Unlearningcited by 0
  23. 2026IP Protection in the Era of Visual Generative AI: A Surveycited by 0
  24. 2026Data Unlearning via Inverse Distillationcited by 0
  25. 2026Weeding Out Bad Seeds: Initial-Noise-Robust Unlearning for Text-to-Image Diffusion Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025When Are Concepts Erased From Diffusion Models?
  2. 2025SPEED: Scalable, Precise, and Efficient Concept Erasure for Diffusion Models
  3. 2026Erasing Without Collateral Damage: Precise Concept Removal in Diffusion Models
  4. 2026Mass Concept Erasure in Diffusion Models with Concept Hierarchy
  5. 2026Erasing Thousands of Concepts: Towards Scalable and Practical Concept Erasure for Text-to-Image Diffusion Models
  6. 2025Revoking Amnesia: RL-based Trajectory Optimization to Resurrect Erased Concepts in Diffusion Models
  7. 2026LURE: Latent Space Unblocking for Multi-Concept Reawakening in Diffusion Models
  8. 2025GrOCE:Graph-Guided Online Concept Erasure for Text-to-Image Diffusion Models