Research map / Image generation / Text-to-image diffusion
Holistic Unlearning Benchmark: A Multi-Faceted Evaluation for Text-to-Image Diffusion Model Unlearning
Saemi Moon, Minjong Lee, Sangdon Park, Dongwoo Kim
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 |
|---|---|---|---|
| unlearn with RECELER | SD v1.5 · HUB overall | target proportion: 0.649 -> 0.086 | Improved |
| unlearn with SA | SD v1.5 · HUB overall | FID: 13.203 -> 32.572 | Hurt |
| unlearn with UCE | SD v1.5 · HUB NSFW | target proportion: 0.647 -> 0.603 | No clear effect |
| unlearn with UCE | SD v1.5 · HUB overall | ImageReward: 0.172 -> 0.193 | Improved |
| unlearn with MACE | SD v1.5 · HUB NSFW | ImageReward: 0.172 -> -1.403 | Hurt |
| unlearn with RECELER | SD v1.5 · HUB overall | multilingual robustness (target proportion): 0.489 -> 0.030 | Improved |
| unlearn with AC | SD v1.5 · HUB overall | target proportion: 0.649 -> 0.301 | Mixed |
| use HUB NSFW prompts instead of I2P prompts | original SD v1.5 · NSFW | target proportion: 0.339 -> 0.647 | Improved |
| unlearn with UCE | SD v1.5 · HUB NSFW | attack robustness: 0.796 -> 0.780 | No clear effect |
Closest papers by meaning
- The Illusion of Unlearning: The Unstable Nature of Machine Unlearning in Text-to-Image Diffusion Models (CVPR 2025 · similarity 0.94)
- UnlearnCanvas: Stylized Image Dataset for Enhanced Machine Unlearning Evaluation in Diffusion Models (NeurIPS 2024 · similarity 0.92)
- Enabling Preference-driven Unlearning in Few-step Distilled Text-to-Image Diffusion Models (NeurIPS 2026 · similarity 0.90)
- Boosting Alignment for Post-Unlearning Text-to-Image Generative Models (NeurIPS 2024 · similarity 0.90)
- Defensive Unlearning with Adversarial Training for Robust Concept Erasure in Diffusion Models (NeurIPS 2024 · similarity 0.90)
- Unlearning Concepts in Diffusion Model via Concept Domain Correction and Concept Preserving Gradient (AAAI 2025 · similarity 0.89)
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.
