统计与数据科学系系列学术报告之五百一十六期

时间：2026年7月14日（星期二）10:00-11:00

主持人：复旦大学管理学院统计与数据科学系郁文教授

地点：史带楼303室

报告人：Prof. Yao Li

The Department of Statistics & Operations research

University of North Carolina at Chapel Hill

题目：Partition-Losses Fine-Tuning: Contamination-Robust Backdoor Unlearning

摘要：The widespread use of large-scale, weakly curated training data and third-party checkpoints makes training convenient, but leaves room for poisoning-based backdoor attacks. These attacks embed a backdoor through data poisoning in the training set: the infected model behaves normally on clean inputs but predicts an attacker-chosen label whenever the trigger appears, posing risks for security-sensitive deployment and model reuse. Post-training fine-tuning has become a practical default defense as it is computationally efficient and does not require control over the original training pipeline. However, most existing post-training fine-tuning defenses rely exclusively on a clean dataset and discard or avoid suspicious inputs, leaving potentially useful information unexploited. As a result, such clean-only approaches optimize benign performance alone and do not directly address the trigger-to-target association. In this paper, we propose a simple and architecture-agnostic post-training method, called Partition-Losses Fine-Tuning (PL), that leverages both clean data and previously overlooked suspicious samples. PL simultaneously encourages correct predictions on clean inputs while discouraging attacker-specified predictions on suspicious inputs, directly breaking the trigger-to-target association. Comprehensive experiments show that PL matches or surpasses clean-only fine-tuning methods under the same computational budget while substantially reducing clean-data requirements, and remains effective under realistic contamination, hyperparameter variation, and cross-attack settings.

个人简介：Yao is an assistant professor of Statistics at UNC Chapel Hill. She was a Ph.D. student at UC Davis working with Prof. Cho-Jui Hsieh and Prof. Thomas C.M. Lee. Her research focuses on developing new algorithms to resolve the real-world difficulties in the machine learning pipeline. She studies both statistical and computational aspects of machine learning models. Currently, she is working on topics related to security of deep learning and computational pathology..