AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

Supercharged One-step Text-to-Image Diffusion Models with Negative Prompts

Viet Nguyen, Anh Nguyen, Trung Dao, Khoi Nguyen, Cuong Pham, Toan Tran, Anh Tran

ICCV 2025 · Image generation › Text-to-image diffusion · approach: Text-driven synthesis and semantic guidance · 7 citations (OpenAlex)

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 changedWhereMeasuredEffect
apply standard CFG with a negative prompt to a one-step modelFLUX.1-schnell 1 step · unwanted feature removalsuccess rate: 44% -> 0%Hurt
apply NASA-I negative prompting training-freeFLUX.1-schnell 1 step · unwanted feature removalsuccess rate: 44% -> 99%Improved
apply CFG=2.5 to DMD one-stepDMD (PixArt-alpha) · HPSv2HPSv2 average: 28.71 -> 25.10Hurt
add NASA-I at inferenceSBv2* (PixArt-alpha) · HPSv2HPSv2 average: 30.34 -> 30.98Improved
add NASA-I at inferenceSBv2*, DMD2, InstaFlow (SD1.5) · HPSv2HPSv2 average: 27.02 -> 27.03No clear effect
add NASA-T distillation with negative promptSBv2* (PixArt-alpha) · HPSv2HPSv2 average: 30.34 -> 30.61Improved
add NASA-T on SD2.1 backboneSBv2* (SD2.1) · HPSv2HPSv2 average: 27.35 -> 27.84Improved
add NASA-T and NASA-I together with alpha=1SBv2* (PixArt-alpha) · HPSv2HPSv2 average: 30.34 -> 31.21Improved
add CFG=1.5 on top of NASA-T model (ablation)NASA-T (PixArt-alpha) · HPSv2HPSv2 average: 30.61 -> 27.97Hurt
put negative prompts into teacher or LoRA teacher during distillation without NASA in student (ablation)SBv2* distillation (PixArt-alpha) · HPSv2HPSv2 average: 30.34 -> 29.84Hurt
add negative prompt to LoRA teacher in addition to teacher (ablation)NASA-T distillation · HPSv2HPSv2 average: 30.61 -> 29.89Hurt
set NASA scale alpha too high (ablation)NASA-I · HPSv2HPSv2Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy