Learning Debuggable Models Through Multi-Objective NAS

Zachariah J Carmichael, Tim Moon, Sam Ade Jacobs

Keywords: Explainable AI, Neural Architecture Search, Trustworthy AI

TL;DR: We develop a proof-of-concept framework for automatically discovering accurate and debuggable architectures.

Abstract: Monumental advances in deep learning have led to unprecedented achievements across various domains. While the performance of deep neural networks is indubitable, the architectural design and interpretability of such models are nontrivial. Research has been introduced to automate the design of neural network architectures through neural architecture search (NAS). Recent progress has made these methods more pragmatic by exploiting distributed computation and novel optimization algorithms. However, there is little work in optimizing architectures for interpretability. To this end, we propose a multi-objective distributed NAS framework that optimizes for both task performance and "introspectability," a surrogate metric for the debuggability of a model. We leverage the non-dominated sorting genetic algorithm (NSGA-II) and explainable AI (XAI) techniques to reward architectures that can be better comprehended by domain experts. The framework is evaluated on several image classification datasets. We demonstrate that jointly optimizing for task error and introspectability leads to more disentangled and debuggable architectures that perform within tolerable error.

Submission Checklist: Yes

Broader Impact Statement: Yes

Paper Availability And License: Yes

Code Of Conduct: Yes

Reviewers: Yes

CPU Hours: 0

GPU Hours: 720

TPU Hours: 0

Evaluation Metrics: Yes