Ruyi Ding (Northeastern University), Tong Zhou (Northeastern University), Lili Su (Northeastern University), Aidong Adam Ding (Northeastern University), Xiaolin Xu (Northeastern University), Yunsi Fei (Northeastern University)

Adapting pre-trained deep learning models to customized tasks has become a popular choice for developers to cope with limited computational resources and data volume. More specifically, probing--training a classifier on a pre-trained encoder--has been widely adopted in transfer learning, which helps to prevent overfitting and catastrophic forgetting. However, such generalizability of pre-trained encoders raises concerns about the potential misuse of probing for harmful applications, such as discriminatory speculation and warfare applications. In this work, we introduce EncoderLock, a novel applicability authorization method designed to protect pre-trained encoders from malicious probing, i.e., yielding poor performance on specified prohibited domains while maintaining their utility in authorized ones. Achieving this balance is challenging because of the opposite optimization objectives and the variety of downstream heads that adversaries can utilize adaptively. To address these challenges, EncoderLock employs two techniques: domain-aware weight selection and updating to restrict applications on prohibited domains/tasks, and self-challenging training scheme that iteratively strengthens resistance against any potential downstream classifiers that adversaries may apply. Moreover, recognizing the potential lack of data from prohibited domains in practical scenarios, we introduce three EncoderLock variants with different levels of data accessibility: supervised (prohibited domain data with labels), unsupervised (prohibited domain data without labels), and zero-shot (no data or labels available). Extensive experiments across fifteen domains and three model architectures demonstrate EncoderLock's effectiveness over baseline methods using non-transferable learning. Additionally, we verify EncoderLock's effectiveness and practicality with a real-world pre-trained Vision Transformer (ViT) encoder from Facebook. These results underscore the valuable contributions EncoderLock brings to the development of responsible AI.

View More Papers

Decoupling Permission Management from Cryptography for Privacy-Preserving Systems

Ruben De Smet (Department of Engineering Technology (INDI), Department of Electronics and Informatics (ETRO), Vrije Universiteit Brussel), Tom Godden (Department of Engineering Technology (INDI), Vrije Universiteit Brussel), Kris Steenhaut (Department of Engineering Technology (INDI), Department of Electronics and Informatics (ETRO), Vrije Universiteit Brussel), An Braeken (Department of Engineering Technology (INDI), Vrije Universiteit Brussel)

Read More

Evaluating the Strength and Availability of Multilingual Passphrase Authentication

Chi-en Amy Tai (University of Waterloo), Urs Hengartner (University of Waterloo), Alexander Wong (University of Waterloo)

Read More

AegisSat: A Satellite Cybersecurity Testbed

Roee Idan, Roy Peled, Aviel Ben Siman Tov, Eli Markus, Boris Zadov, Ofir Chodeda, Yohai Fadida (Ben Gurion University of the Negev), Oliver Holschke, Jan Plachy (T-Labs (Research & Innovation)), Yuval Elovici, Asaf Shabtai (Ben Gurion University of the Negev)

Read More

On Borrowed Time – Preventing Static Side-Channel Analysis

Robert Dumitru (Ruhr University Bochum and The University of Adelaide), Thorben Moos (UCLouvain), Andrew Wabnitz (Defence Science and Technology Group), Yuval Yarom (Ruhr University Bochum)

Read More