Awesome Concept Erasing

Table of Contents

• Awesome Concept Erasing
  • Table of Contents
  • Surveys
  • Concept Erasing
    • Model Fintuning
  • Video Concept Erasing
  • Model Editing
  • Benchmark

Surveys

• A Comprehensive Survey on Concept Erasure in Text-to-Image Diffusion Models (Feb, 2025)

Concept Erasing

Model Fintuning

• Fantastic Targets for Concept Erasure in Diffusion Models and Where To Find Them (2025, ICLR 2025)
  

• SAeUron: Interpretable Concept Unlearning in Diffusion Models with Sparse Autoencoders (2025, arXiv)
  

• Get What You Want, Not What You Don't: Image Content Suppression for Text-to-Image Diffusion Models (2024, ICLR 2024)
  

• Erasing Undesirable Concepts in Diffusion Models with Adversarial Preservation (2024, NeurIPS 2024)
  

• Reliable and Efficient Concept Erasure of Text-to-Image Diffusion Models (2024, ECCV 2024)
  

• R.A.C.E.: Robust Adversarial Concept Erasure for Secure Text-to-Image Diffusion Model (2024, ECCV 2024)
  

• Receler: Reliable Concept Erasing of Text-to-Image Diffusion Models via Lightweight Erasers (2023, ECCV 2024)
  

• [Latent guard: a safety framework for text-to-image generation](https://arxiv.org/abs/2311.17717) (2023, ECCV 2024)
  

• MACE: Mass Concept Erasure in Diffusion Models (2024, CVPR 2024)
  

• One-Dimensional Adapter to Rule Them All: Concepts, Diffusion Models and Erasing Applications (2024, CVPR 2024)
  

• SafeGen: Mitigating Sexually Explicit Content Generation in Text-to-Image Models (2024, ACM CCS 2024)
  

• ConceptPrune: Concept Editing in Diffusion Models via Skilled Neuron Pruning (2024, arXiv)
  

• Pruning for Robust Concept Erasing in Diffusion Models (2024, arXiv)
  

• Defensive Unlearning with Adversarial Training for Robust Concept Erasure in Diffusion Models (2024, arXiv)
  

• Unlearning Concepts in Diffusion Model via Concept Domain Correction and Concept Preserving Gradient (2024, arXiv)
  

• Robust Concept Erasure Using Task Vectors (2024, arXiv)
  

• Separable Multi-Concept Erasure from Diffusion Models (2024, arXiv)
  

• EraseDiff: Erasing Data Influence in Diffusion Models (2024, arXiv)
  

• All but One: Surgical Concept Erasing with Model Preservation in Text-to-Image Diffusion Models (2023, AAAI 2024)
  

• SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation (2023, ICLR 2024)
  

• Unified Concept Editing in Diffusion Models (2023, WACV 2024)
  

• Selective Amnesia: A Continual Learning Approach to Forgetting in Deep Generative Models (2023, NeurIPS 2023)
  

• Ablating Concepts in Text-to-Image Diffusion Models (2023, ICCV 2023)
  

• Erasing Concepts from Diffusion Models (2023, ICCV 2023)
  

• Degeneration-Tuning: Using Scrambled Grid shield Unwanted Concepts from Stable Diffusion (2023, ACM MM 2023)
  

• Implicit Concept Removal of Diffusion Models (2023, arXiv)
  

• Forget-Me-Not: Learning to Forget in Text-to-Image Diffusion Models (2023, arXiv)
  

• Probing Unlearned Diffusion Models: A Transferable Adversarial Attack Perspective (2024, arXiv)
  

• Circumventing Concept Erasure Methods For Text-to-Image Generative Models (2023, ICLR 2024)
  

• Ring-A-Bell! How Reliable are Concept Removal Methods for Diffusion Models? (2023, ICLR 2024)
  

• To Generate or Not? Safety-Driven Unlearned Diffusion Models Are Still Easy To Generate Unsafe Images ... For Now (2023, ECCV 2024)
  

• Riatig: Reliable and imperceptible adversarial text-to-image generation with natural prompts (2023, CVPR 2023)
  

Video Concept Erasing

• Unlearning Concepts from Text-to-Video Diffusion Models (Jul, 2024)

  

Model Editing

• ReFACT: Updating Text-to-Image Models by Editing the Text Encoder (May, 2024) (NAACL 2024)

  

• Towards Memorization-Free Diffusion Models (Apr, 2024) (CVPR 2024)

  

Benchmark

• UnlearnCanvas: A Stylized Image Dataset to Benchmark Machine Unlearning for Diffusion Models (2024, arXiv)
  
• T2VSafetyBench: Evaluating the Safety of Text-to-Video Generative Models (Sep,2024, NeurIPS 2024)
  
• SafeSora: Towards Safety Alignment of Text2Video Generation via a Human Preference Dataset (Jun,2024, NeurIPS 2024)