We present the Membership Inference Test (MINT) Demo 2, a framework designed to improve transparency in machine learning training processes. MINT is a technique for experimentally determining whether specific data were used during machine learning model training. We establish the theoretical framework and propose multiple architectures for MINT depending on the amount of information known about the models being audited.
Experimental results using a popular face recognition model, four state-of-the-art LLMs, and multiple diverse large-scale public image and text databases achieve promising accuracy levels in the detection of training data, reaching up to 90%. Building on these results, we introduce a comprehensive web platform that expands these capabilities to image and text modalities. The platform integrates a diverse technological stack, including MINT, aMINT, and gMINT, allowing users to audit a wide range of models. This demonstrator aims to promote AI transparency and provides a practical tool to foster compliance with emerging AI regulations.
Blogger's Review: The introduction of MINT is a significant step towards enhancing AI transparency. It not only improves the auditability of model training data but also provides robust support for regulatory compliance. As AI technology rapidly evolves, tools like this will become increasingly important and worthy of attention.