ABLE: Using Adversarial Pairs to Construct Local Models for Explaining Model Predictions

ABLE introduces a robust framework for model interpretability by leveraging adversarial pairs to construct stable local models, offering superior fidelity ov...

Level: advanced

By Krishna Khadka, Sunny Shree, Pujan Budhathoki, Yu Lei, Raghu Kacker, D. Richard Kuhn

Category: research