Research map / Image generation / Text-to-image diffusion
GenErase: Generalizable and Semantically-Aware Concept Erasure in Diffusion Models
Korada Sri Vardhana, Soma Biswas
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 per-token preserve projector (Safe Semantic Subspace) (ablation) | GenErase · Mark Zuckerberg erasure (qualitative) | non-target fidelity | Improved |
| increase number of erased identities from 1 to 50 (ablation) | GenErase · Scalability (Table 3) | HM: 42.29 -> 41.63 | Hurt |
| use GenErase instead of AdaVD for single-identity erasure | Stable Diffusion v1.4 · Single-celebrity erasure | ESR: 78.33 -> 81.83 | Improved |
| use GenErase instead of AdaVD for single-identity erasure | Stable Diffusion v1.4 · Single-celebrity erasure | non-target FID: 8.09 -> 11.85 | Hurt |
| use GenErase instead of AdaVD on paraphrase benchmark | Stable Diffusion v1.4 · GenBench-40 (average) | failure rate (%): 29.88 -> 22.70 | Improved |
| use GenErase instead of AdaVD for object erasure | Stable Diffusion v1.4 · Object erasure | non-target FID: 12.63 -> 15.43 | Hurt |
| lower gate threshold to 0.05 for nudity suppression | GenErase · I2P nudity | success rate (%): ~83% -> 87.35% | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SAFREE — SAFREE: Training-Free and Adaptive Guard for Safe Text-to-Image And Video Generation (ICLR 2025) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- MACE: Mass Concept Erasure in Diffusion Models (CVPR 2024 · similarity 0.93)
- Semantic Surgery: Zero-Shot Concept Erasure in Diffusion Models (NeurIPS 2025 · similarity 0.93)
- Growth Inhibitors for Suppressing Inappropriate Image Concepts in Diffusion Models (ICLR 2025 · similarity 0.92)
- Precise, Fast, and Low-cost Concept Erasure in Value Space: Orthogonal Complement Matters (CVPR 2025 · similarity 0.92)
- Localized Concept Erasure for Text-to-Image Diffusion Models Using Training-Free Gated Low-Rank Adaptation (CVPR 2025 · similarity 0.91)
- Differential Vector Erasure: Unified Training-Free Concept Erasure for Flow Matching Models (NeurIPS 2026 · similarity 0.91)
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.
