Benchmarking Unlearning for Vision Transformers

TL;DR

This paper establishes a comprehensive benchmark for machine unlearning in Vision Transformers, comparing various algorithms across datasets and protocols to evaluate their effectiveness and memorization characteristics.

Contribution

It is the first to benchmark MU algorithms on Vision Transformers, analyzing their performance, memorization, and impact of dataset complexity across multiple protocols.

Findings

MU algorithms vary significantly in VT settings

Memorization influences unlearning effectiveness

Benchmark provides a baseline for future research

Abstract

Research in machine unlearning (MU) has gained strong momentum: MU is now widely regarded as a critical capability for building safe and fair AI. In parallel, research into transformer architectures for computer vision tasks has been highly successful: Increasingly, Vision Transformers (VTs) emerge as strong alternatives to CNNs. Yet, MU research for vision tasks has largely centered on CNNs, not VTs. While benchmarking MU efforts have addressed LLMs, diffusion models, and CNNs, none exist for VTs. This work is the first to attempt this, benchmarking MU algorithm performance in different VT families (ViT and Swin-T) and at different capacities. The work employs (i) different datasets, selected to assess the impacts of dataset scale and complexity; (ii) different MU algorithms, selected to represent fundamentally different approaches for MU; and (iii) both single-shot and continual unlearning protocols. Additionally, it focuses on benchmarking MU algorithms that leverage training data memorization, since leveraging memorization has been recently discovered to significantly improve the performance of previously SOTA algorithms. En route, the work characterizes how VTs memorize training data relative to CNNs, and assesses the impact of different memorization proxies on performance. The benchmark uses unified evaluation metrics that capture two complementary notions of forget quality along with accuracy on unseen (test) data and on retained data. Overall, this work offers a benchmarking basis, enabling reproducible, fair, and comprehensive comparisons of existing (and future) MU algorithms on VTs. And, for the first time, it sheds light on how well existing algorithms work in VT settings, establishing a promising reference performance baseline.