Research map / Image generation / Text-to-image diffusion
Translation of Text Embedding via Delta Vector to Suppress Strongly Entangled Content in Text-to-Image Diffusion Models
Eunseo Koh, Seunghoo Hong, Tae-Young Kim, Simon S. Woo, Jae-Pil Heo
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 |
|---|---|---|---|
| apply delta to both key (positive) and value (negative) instead of key only (ablation) | SSDV · SEP-Benchmark (SD) | IFID: 27.49 -> 72.63 | Improved |
| apply delta to both key (positive) and value (negative) instead of value only (ablation) | SSDV · SEP-Benchmark (SD) | IFID: 23.01 -> 72.63 | Improved |
| apply delta to key and value on OwlViT detection score (ablation) | SSDV · SEP-Benchmark (SD) | OwlViT detection score: 0.049 -> 0.012 | Improved |
| subtract negative-content embedding (delta vector) from word embedding with selective key/value use, versus prior suppression methods | Stable Diffusion · SEP-Benchmark (SD) | DetScore: 0.220 -> 0.113 | Improved |
| use optimized delta instead of zero-shot delta on DreamBooth models (ablation) | SSDV on DreamBooth · DreamBooth personalized models | human preference (%): 26.8 -> 73.2 | Improved |
| use SSDV instead of Inst-Inpaint on S* toy duck yellow | DreamBooth S* toy duck · DreamBooth S* toy duck - yellow | IFID: 279.61 -> 118.26 | Hurt |
| use negative prompt for suppression of entangled content | Stable Diffusion · SEP-Benchmark (SD) | CLIP score: 17.60 -> 16.46 | Mixed |
Closest papers by meaning
- Get What You Want, Not What You Don't: Image Content Suppression for Text-to-Image Diffusion Models (ICLR 2024 · similarity 0.91)
- Magnet: We Never Know How Text-to-Image Diffusion Models Work, Until We Learn How Vision-Language Models Function (NeurIPS 2024 · similarity 0.87)
- Diffusion Bridge: Leveraging Diffusion Model to Reduce the Modality Gap Between Text and Vision for Zero-Shot Image Captioning (CVPR 2025 · similarity 0.87)
- A Cat Is A Cat (Not A Dog!): Unraveling Information Mix-ups in Text-to-Image Encoders through Causal Analysis and Embedding Optimization (NeurIPS 2024 · similarity 0.86)
- Less is More: Masking Elements in Image Condition Features Avoids Content Leakages in Style Transfer Diffusion Models (ICLR 2025 · similarity 0.86)
- LightFair: Towards an Efficient Alternative for Fair T2I Diffusion via Debiasing Pre-trained Text Encoders (NeurIPS 2025 · similarity 0.86)
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.
