6 papers
GRACE: Adaptive Concept Erasure with Geometry-Guided Retention in Diffusion Models
Qinghui Gong, Yihuai Liang, Yuanlun Xie +3
Text-to-image (T2I) diffusion models inevitably internalize sensitive or non-compliant concepts from large-scale pretraining data, necessitating post-hoc concept erasure. However,…
Gaussian Core LoRA: Distribution-Aware Dynamic Adaptation for Broad Concept Erasure
Qinghui Gong, Xunlei Chen, Yu-Xuan Zhang +2
Concept erasure aims to suppress unsafe, privacy-sensitive, or undesirable generations in text-to-image diffusion models while preserving benign semantics, visual quality, and depl…
STU: Stateful Test-Time Unlearning via Restricted Knowledge Boundary Control
Xunlei Chen, Qinghui Gong, Ruini Xue +3
Controlling restricted knowledge in large language models is essential for model alignment and safe deployment. Test-time unlearning avoids costly retraining and parameter updates…
DSS: Dynamic Semantic Steering for Robust Concept Erasure in Diffusion Models
Qinghui Gong, Zhengchun Zhou, Hua Meng +2
Text-to-image (T2I) diffusion models have introduced new security risks, as adversaries can exploit flexible text prompts to induce the generation of sensitive or policy-violating…
Understand Then Memory: A Cognitive Gist-Driven RAG Framework with Global Semantic Diffusion
Pengcheng Zhou, Haochen Li, Zhiqiang Nie +4
Retrieval-Augmented Generation (RAG) effectively mitigates hallucinations in LLMs by incorporating external knowledge. However, the inherent discrete representation of text in exis…
Orthogonal Soft Pruning for Efficient Class Unlearning
Qinghui Gong, Xue Yang, Xiaohu Tang
Efficient and controllable data unlearning in federated learning remains challenging, due to the trade-off between forgetting and retention performance. Especially under non-indepe…