EraseSAE introduces a surgical concept erasure method for text-to-video diffusion models, using sparse autoencoders to decompose activations into interpretable, monosemantic features. The framework employs a contrastive attribution mechanism to isolate concept-specific kernels and applies timestep-resolved masks during inference to remove target concepts while preserving unrelated content. Experiments show that EraseSAE achieves precise, robust concept removal with minimal quality loss, outperforming existing methods.
arXiv:2606. 31699v1 Announce Type: cross Abstract: Sparse autoencoders (SAEs) have recently been proposed as interpretable tools for concept-level manipulation, under the assumption that isolated features can serve as controllable intervention points.
By Enrico Cassano, Riccardo Renzulli, Rayyan Ahmed, Marco Grangetto, Stephan Alaniz
arXiv:2609.09909v1 Announce Type: new
Abstract: Although text-to-image diffusion models generally exhibit strong prompt-following ability, we identify a persistent and previously underexplored failur...
By Yifan Yuan, Xiangyu Liu, Hongming Shan, Yu Han, Yu Jiang, Hao Tan, Junping Zhang, Linlin Shen
arXiv:2607. 23492v1 Announce Type: cross Abstract: Concept erasure techniques (CETs) edit text-to-image diffusion models to erase undesired targets such as NSFW content or copyrighted styles, while preserving model utility on benign concepts.
By Shaswati Saha, Rajasekhar Anguluri, Manas Gaur
arXiv:2607. 06432v1 Announce Type: cross Abstract: Concept unlearning in text-to-image diffusion models is critical for safe and practical deployment: with rising privacy concerns, copyright disputes, trademark constraints, and safety regulations, deployed systems must be able to suppress unwanted concepts after training.
By Naveen George, Naoki Murata, Yuhta Takida, Konda Reddy Mopuri, Yuki Mitsufuji
Training-free concept erasure is an attractive mechanism for controlling text-to-image diffusion models, but precise erasure often comes at the cost of damaging semantically related non-target concepts. Existing value-space methods remove the component of each cross-attention value along the target concept direction, implicitly treating target identity and shared visual structure as the same signal.
TINA+ is a diffusion-consistent, text‑free inversion attack that probes residual visual knowledge in diffusion models after concept erasure. By using optimization‑based inversion and diffusion‑consistent trajectory regularization, it suppresses spurious trajectories that could falsely indicate retained concepts. Experiments across multiple erasure methods, tasks, and model architectures show that TINA+ reliably recovers erased concepts, revealing that many current techniques only sever text‑image links rather than eliminating underlying visual knowledge.
arXiv:2607. 08605v1 Announce Type: cross Abstract: Sparse autoencoders (SAEs) have emerged as a promising technique for mechanistic interpretability by learning a set of sparse latent features in large models, each of which encodes a distinct concept.
By Weiduo Liao, Yunqiao Yang, Ying Wei
arXiv:2609.01433v1 Announce Type: new
Abstract: Concept erasure aims to suppress unsafe, privacy-sensitive, or undesirable generations in text-to-image diffusion models while preserving benign semant...
By Qinghui Gong, Xunlei Chen, Yu-Xuan Zhang, Hua Meng, Zhengchun Zhou
arXiv:2606.03695v2 Announce Type: replace
Abstract: As language models are increasingly deployed in real-world applications, the ability to erase specific knowledge from them becomes critical for saf...
By Clara Haya Suslik, Or Shafran, Mor Geva
arXiv:2608.23864v1 Announce Type: new
Abstract: Visual tokenizers increasingly inject semantic supervision into latent spaces to make downstream diffusion easier. Yet how these semantics should be or...
By Junqiu Yu, Pandeng Li, Yikai Wang, Jiaxing Zhao, Yujie Wei, Kaixun Jiang, Quanhao Li, Hongtao Yu, Zhihang Liu, Zhaohe Liao, Junjie Zhou, Yun Zheng, Yu Liu, Yanwei Fu
arXiv:2606. 25548v1 Announce Type: cross Abstract: Image generative models are trained on massive, largely uncurated internet-scale datasets that contain undesirable visual concepts.
By Aditya Kumar, Pierre Joly, Adam Dziedzic, Franziska Boenisch