arXiv AI

Concept Unlearning via Cross-Attention Activation Projection for Diffusion Models

Hugging Face Trending Papers
Jul 6

Erasing Without Collateral Damage: Precise Concept Removal in Diffusion Models

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.

arXiv AI
Oct 15

No Concept Escapes the Audit: Auditing-Aware Unlearning for Verifiable Concept Erasure in Diffusion Models

The paper introduces AVCE, a method for verifiable concept erasure in diffusion models that audits latent representations to identify vulnerable directions and then edits cross‑attention and self‑attention projections in closed form. By fine‑tuning with pathway‑level auditing losses and orthogonal gradient projection, AVCE reduces attack success rates by 5.07× and improves auditing scores by 3.84× compared to the best baseline while maintaining generation quality.

By Kaiyuan Deng, Yuchen Li, Gen Li, Yang Xiao, Geng Yuan, Xiaoyong Yuan, Bo Hui, Xiaolong Ma
arXiv Machine Learning
Oct 23

Disentangled Sparse Representations for Concept-Separated Diffusion Unlearning

The paper introduces SAEParate, a sparse autoencoder approach that explicitly separates overlapping concepts in the latent space of text-to-image diffusion models. It employs concept-aware contrastive learning and cross-attention-guided spatial aggregation to focus on concept-relevant regions, while a nonlinear encoder transformation enhances expressive capacity. Experiments on UnlearnCanvas show state‑of‑the‑art performance, especially in challenging joint style‑object unlearning scenarios where prior methods struggle with interference between concepts.

By Hyeonjin Kim, Hangyeol Jung, Heechan Yun, Sungjun Yun, Dong-Jun Han
arXiv Computer Vision
Oct 2

RASteer: Retain-Aware Activation Steering for Concept Erasure in Diffusion Models

RASteer is a training‑free method for concept erasure in text‑to‑image diffusion models that retains other concepts. It constructs a retain subspace from concepts to preserve, then removes components aligned with this subspace from the erasure direction using Retain‑Orthogonal Steering (ROS). Overlap‑Adaptive Calibration (OAC) further adjusts the removal of shared components at each layer and denoising step, balancing target erasure with concept preservation. Experiments show RASteer matches or outperforms existing activation steering and weight editing baselines on unsafe‑content, instance, and artistic‑style erasure tasks across multiple backbones and benchmarks.

By Yongliang Wu, Haori Lu, Yulun Wu, Jinqi Luo, Xingyu Zhu, Yaoyao Liu
arXiv AI
Jul 8

TILDE: TILt-based Distributional Erasure for Concept Unlearning

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
arXiv Machine Learning
Oct 8

ORCA: Hunting Compositional Failures in Text-to-Image Diffusion

ORCA (Orthogonal Residual Compositional Alignment) is a method that improves text-to-image diffusion models by aligning the diffusion transformer’s latent space with a low‑rank target derived from a frozen visual encoder. It introduces an auxiliary loss that uses a predictor with an orthogonal basis parameterised by a learned residual between T5 and CLIP embeddings, providing a prompt‑dependent signal for selecting the visual readout subspace. Experiments on three diffusion‑transformer backbones show that ORCA improves FID and GenEval scores, especially on attribute binding, spatial relations, and multi‑object prompts, without adding inference‑time cost.

By Arshia Hemmat, Amirhossein Vahidi, Amitis Shidani, Mohammad Vali Sanian, Hesam Asadollahzadeh, Aryan Yazdan Parast, Mohammad Lotfollahi
arXiv Computer Vision
Sep 30

Weeding Out Bad Seeds: Initial-Noise-Robust Unlearning for Text-to-Image Diffusion Models

arXiv:2609.37537v1 Announce Type: new Abstract: Machine unlearning has emerged as a critical post-hoc safety measure to erase sensitive concepts from Text-to-Image (T2I) models without prohibitive re...

By Arian Komaei Koma, Seyed Amir Kasaei, Aida Aryafar, Matin Ghiasi, Ali Aghayari, Amirhossein Souri, Mohammad Mosayyebi, AmirMahdi Sadeghzadeh, Mohammad Hossein Rohban
Hugging Face Trending Papers
Aug 18

TINA+: Probing Residual Visual Knowledge in Unlearned Diffusion Models via Diffusion-Consistent Text-Free Inversion

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 AI
Nov 5

ReMoA: Recursive Modular Adaptation via Sparse Latent Refinement in Multimodal Diffusion Models

ReMoA introduces a Recursive Modular Adaptation framework that embeds sparse neural modules into joint attention layers of diffusion models, enabling iterative refinement of visual tokens across multiple recursion steps. The method leverages modular human cognition principles to share parameters efficiently while improving text-to-image generation. Experiments on ImageNet and GenEval/DPG benchmarks show that ReMoA enhances object detail, alignment, and recall without adding extra parameters.

By Yuwei Sun, Yuxuan Yao, Hui Li, Weihao Yuan, Siyu Zhu