Research map / Image generation / Text-to-image diffusion
RAIGen: Rare Attribute Identification in Text-to-Image Generative Models
Silpa Vadakkeeveetil Sreelatha, Dan Wang, Serge Belongie, Muhammad Awais, Anjan Dutta
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 |
|---|---|---|---|
| RAIGen-guided prompt revision | SD v1.4 · WinoBias | deviation ratio: 0.50 -> 0.22 | Improved |
| RAIGen-guided prompt revision | SDXL · WinoBias | deviation ratio: 0.49 -> 0.23 | Improved |
| RAIGen-guided prompt revision | SD v1.4 · COCO | deviation ratio: 0.53 -> 0.25 | Improved |
| RAIGen-guided prompt revision | SDXL · COCO | deviation ratio: 0.51 -> 0.28 | Improved |
| RAIGen-guided prompt revision | SD v1.4 · WinoBias | CLIP alignment: 20.30 -> 19.80 | Hurt |
| RAIGen-guided prompt revision | SDXL · COCO | CLIP alignment: 20.52 -> 19.83 | Hurt |
| RAIGen-guided prompt revision | SD v1.4 · WinoBias | NLL under base prompt: 1.917 -> 1.935 | Improved |
| scale from SD v1.4 to SDXL | RAIGen minority attributes · WinoBias | attribute presence: 0.205 -> 0.194 | No clear effect |
| suppress dominant White attribute with negative prompt (ablation) | SDXL · Doctor/Manager | attribute prevalence | Mixed |
| use standard SAE instead of Matryoshka SAE (ablation) | RAIGen · qualitative | interpretability | Hurt |
| use activation frequency alone as minority score (ablation) | RAIGen · toy data | Spearman correlation: rho approx 0.991 | Improved |
Closest papers by meaning
- Attribute Based Interpretable Evaluation Metrics for Generative Models (ICML 2024 · similarity 0.89)
- Balancing Act: Distribution-Guided Debiasing in Diffusion Models (CVPR 2024 · similarity 0.88)
- FairImagen: Post-Processing for Bias Mitigation in Text-to-Image Models (NeurIPS 2025 · similarity 0.88)
- Localizing and Editing Knowledge In Text-to-Image Generative Models (ICLR 2024 · similarity 0.87)
- Rethinking Training for De-biasing Text-to-Image Generation: Unlocking the Potential of Stable Diffusion (CVPR 2025 · similarity 0.87)
- Continuous, Subject-Specific Attribute Control in T2I Models by Identifying Semantic Directions (CVPR 2025 · similarity 0.87)
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.
