Massif: Interactive Interpretation of Adversarial Attacks on Deep Learning
This addresses the need for interpretability in high-stakes applications like autonomous cars and healthcare, though it is incremental as it builds on existing visualization methods.
The paper tackles the problem of deep neural networks being vulnerable to adversarial attacks by developing Massif, an interactive tool that identifies and visualizes affected neurons, helping users understand vulnerable input features.
Deep neural networks (DNNs) are increasingly powering high-stakes applications such as autonomous cars and healthcare; however, DNNs are often treated as "black boxes" in such applications. Recent research has also revealed that DNNs are highly vulnerable to adversarial attacks, raising serious concerns over deploying DNNs in the real world. To overcome these deficiencies, we are developing Massif, an interactive tool for deciphering adversarial attacks. Massif identifies and interactively visualizes neurons and their connections inside a DNN that are strongly activated or suppressed by an adversarial attack. Massif provides both a high-level, interpretable overview of the effect of an attack on a DNN, and a low-level, detailed description of the affected neurons. These tightly coupled views in Massif help people better understand which input features are most vulnerable or important for correct predictions.