Active Membership Inference Test (aMINT): Enhancing Model Auditability with Multi-Task Learning
This work addresses the need for transparency and security in AI deployments, particularly for privacy and copyright protection, though it appears incremental as it builds on existing membership inference techniques.
The paper tackles the problem of detecting whether data was used in training machine learning models by proposing Active MINT, a multi-task learning method that achieves over 80% accuracy in membership inference tests across various neural networks and benchmarks.
Active Membership Inference Test (aMINT) is a method designed to detect whether given data were used during the training of machine learning models. In Active MINT, we propose a novel multitask learning process that involves training simultaneously two models: the original or Audited Model, and a secondary model, referred to as the MINT Model, responsible for identifying the data used for training the Audited Model. This novel multi-task learning approach has been designed to incorporate the auditability of the model as an optimization objective during the training process of neural networks. The proposed approach incorporates intermediate activation maps as inputs to the MINT layers, which are trained to enhance the detection of training data. We present results using a wide range of neural networks, from lighter architectures such as MobileNet to more complex ones such as Vision Transformers, evaluated in 5 public benchmarks. Our proposed Active MINT achieves over 80% accuracy in detecting if given data was used for training, significantly outperforming previous approaches in the literature. Our aMINT and related methodological developments contribute to increasing transparency in AI models, facilitating stronger safeguards in AI deployments to achieve proper security, privacy, and copyright protection.