AI Paper Atlas

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

T2I-RiskyPrompt: A Benchmark for Safety Evaluation, Attack, and Defense on Text-to-Image Model

Chenyu Zhang, Tairen Zhang, Lanjun Wang, Ruidong Chen, Wenhui Li, Anan Liu

AAAI 2026 · Image generation › Text-to-image diffusion · approach: T2I evaluation and controllable adapters

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
use a more capable T2I modelT2I models · T2I-RiskyPromptrisk ratio: SD1.4 0.790 -> SD3 0.923Hurt
apply MACE fine-tuning defenseSD1.4 · T2I-RiskyPrompt pornographyrisk ratio: 0.976 -> 0.190Improved
apply TRCE defenseSD1.4 · T2I-RiskyPrompt pornographyrisk ratio: 0.976 -> 0.032Improved
apply SPEED model editingSD1.4 · T2I-RiskyPrompt copyrightrisk ratio: 0.826 -> 0.854No clear effect
apply SLD inference guidanceSD1.4 · T2I-RiskyPrompt pornographyrisk ratio: 0.976 -> 0.874Mixed
apply stronger defense (MACE)SD1.4 · T2I-RiskyPrompt copyright/politicalCLIP-S: 30.97 -> 23.91Hurt
apply SPEED for political conceptsSD1.4 · T2I-RiskyPrompt politicalrisk ratio: 0.914 -> 0.201Improved
use keyword filterSD1.4 · T2I-RiskyPrompt avgrisk ratio: 0.790 -> 0.119Improved
use NSFW-I image filterSD1.4 · T2I-RiskyPrompt avgrisk ratio: 0.790 -> 0.662Improved
ensemble four filtersSD1.4 · T2I-RiskyPrompt avgrisk ratio: 0.119 -> 0.089Improved
apply jailbreak attacks (MMA, MJA) against filtersSD1.4 + filters · T2I-RiskyPromptrisk ratioHurt

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