AI & ML New Capability

Z-Erase introduces the first concept erasure method for single-stream diffusion transformers, preventing generation collapse in new unified architectures.

March 27, 2026

Original Paper

Z-Erase: Enabling Concept Erasure in Single-Stream Diffusion Transformers

Nanxiang Jiang, Zhaoxin Fan, Baisen Wang, Daiheng Gao, Junhang Cheng, Jifeng Guo, Yalan Qin, Yeying Jin, Hongwei Zheng, Faguo Wu, Wenjun Wu

arXiv · 2603.25074

The Takeaway

As T2I models shift from U-Net to single-stream transformers (like Flux or Z-Image), traditional erasure fails. This provides the necessary safety mechanism to remove NSFW or copyrighted concepts from these emerging high-performance architectures.

From the abstract

Concept erasure serves as a vital safety mechanism for removing unwanted concepts from text-to-image (T2I) models. While extensively studied in U-Net and dual-stream architectures (e.g., Flux), this task remains under-explored in the recent emerging paradigm of single-stream diffusion transformers (e.g., Z-Image). In this new paradigm, text and image tokens are processed as a single unified sequence via shared parameters. Consequently, directly applying prior erasure methods typically leads to g