Research map / Adversarial robustness and security / Certified and empirical robustness
AEGIS: Adversarial Target–Guided Retention-Data-Free Robust Concept Erasure from Diffusion Models
Fengpeng Li, Kemou Li, Qizhou Wang, Bo Han, Jiantao Zhou
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add adversarial erasure target (AET) (ablation) | AEGIS · Nudity, UnlearnDiffAtk | ASR: 52.11% -> 8.45% | Improved |
| add parameter regularization instead of COCO retention data (ablation) | AEGIS · Nudity | FID: 18.15 -> 17.43 | Improved |
| add directional gradient rectification (ablation) | AEGIS · Nudity | ASR: 26.24% -> 8.45% | Improved |
| add directional gradient rectification (ablation) | AEGIS · Nudity | FID: 19.84 -> 17.43 | Improved |
| use dynamic omega rather than fixed omega=1 (ablation) | AEGIS · Nudity | ASR: 14.08% -> 8.45% | Improved |
| apply AET to ESD (ablation) | ESD · Nudity, UnlearnDiffAtk | ASR: 76.05% -> 26.06% | Improved |
| apply AET to AdvUnlearn (ablation) | AdvUnlearn · Nudity, UnlearnDiffAtk | ASR: 64.79% -> 5.63% | Improved |
| add GRP to AdvUnlearn (AEGIS w/o AET) (ablation) | AdvUnlearn · Nudity, UnlearnDiffAtk | FID: 19.34 -> 17.54 | Improved |
| move base model from SD v1.4 to SD v2.1 | AEGIS · Van Gogh | ASR1: 36% -> 40% | Hurt |
| AEGIS vs SH on Church | AEGIS · Church | FID: 68.02 -> 19.06 | Improved |
Closest papers by meaning
- Defensive Unlearning with Adversarial Training for Robust Concept Erasure in Diffusion Models (NeurIPS 2024 · similarity 0.92)
- Erasing Undesirable Concepts in Diffusion Models with Adversarial Preservation (NeurIPS 2024 · similarity 0.91)
- When Are Concepts Erased From Diffusion Models? (NeurIPS 2025 · similarity 0.91)
- STEREO: A Two-Stage Framework for Adversarially Robust Concept Erasing from Text-to-Image Diffusion Models (CVPR 2025 · similarity 0.90)
- Concept Pinpoint Eraser for Text-to-image Diffusion Models via Residual Attention Gate (ICLR 2025 · similarity 0.89)
- EraseFlow: Learning Concept Erasure Policies via GFlowNet-Driven Alignment (NeurIPS 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
