Transstratal Adversarial Attack: Compromising Multi-Layered Defenses in Text-to-Image Models

NeurIPSSpotlight2025

Authors
Chunlong Xie, Kangjie Chen, Shangwei Guo, Shudong Zhang, Tianwei Zhang, Tao Xiang
Venue
NeurIPS 2025
Track
Spotlight

TL;DR

Modern Text-to-Image (T2I) models deploy multi-layered defenses to block Not-Safe-For-Work (NSFW) content generation…

Opening excerpt from the authors’ abstract. source

Read the paper

Topics

text-to-image adversarial

Related NeurIPS papers

← All NeurIPS 2025 Spotlight papers · Browse the whole archive