Category: Paper

Our papers “DIFT: Protecting Contrastive Learning against Data Poisoning Backdoor Attacks”, “Class-feature Watermark: A Resilient Black-box Watermark Against Model Extraction Attacks”, “How Much Do Large Language Model Cheat on Evaluation? Benchmarking Overestimation under the One-Time-Pad-Based Framework”, and “Stochastic Universal Adversarial Perturbations with Fixed Optimization Constraint and Ensured High-probability Transferability” have been accepted by AAAI ’26….

Our papers “‘Virus Infection Attack on LLMs: Your Poisoning Can Spread “VIA” Synthetic Data” and “Toward Efficient Inference Attacks: Shadow Model Sharing via Mixture-of-Experts” are accepted by Annual Conference on Neural Information Processing Systems (NeurIPS), 2025. Congratulations to Zi and Li!

Our papers “‘Reminiscence Attack on Residuals: Exploiting Approximate Machine Unlearning for Privacy” and “Federated Continuous Category Discovery and Learning” are accepted by International Conference on Computer Vision (ICCV), 2025. Congratulations to Yaxin and Lixu!

Our paper “PrivAGM: Secure Construction of Differentially Private Directed Attributed Graph Models on Decentralized Social Graphs” is accepted by VLDB 2025.

Our paper “‘Yes, My LoRD.’ Guiding Language Model Extraction with Locality Reinforced Distillation” is accepted by the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025) Main Conference. Congratulations to Zi!

Our paper “Does Low Rank Adaptation Lead to Lower Robustness against Training-Time Attacks?” is accepted by 42nd International Conference on Machine Learning (ICML ’25).

In addition to the previous news that two papers were accepted by ICDE 2025 in round 1, five more papers below are accepted in round 2.

Our paper “Randomized Response to Randomized Index: Answering Subset Counting Queries with Local Differential Privacy” is accepted by IEEE Symposium on Security and Privacy (S&P), 2025.

Our paper “Privacy for Free: Leveraging Local Differential Privacy Perturbed Data from Multiple Services” is accepted by The Proceedings of the VLDB Endowment (PVLDB), 2025.

Our papers “MER-Inspector: Assessing Model Extraction Risks from An Attack-Agnostic Perspective” and “FUNU: Boosting Machine Unlearning Efficiency by Filtering Unnecessary Unlearning” are accepted by ACM TheWebConf 2025 (WWW 2025).