Modeling Latent Attention Within Neural Networks

TL;DR

This paper introduces a visualization method for neural networks that highlights which input features influence decisions, enhancing interpretability across various domains like vision, language, and reinforcement learning.

Contribution

It presents a general, dataset-centric approach to visualize and interpret the internal attention mechanisms of neural networks across multiple modalities.

Findings

Effective visualization of attention masks in diverse neural architectures

Improved understanding of input attribute importance in decision-making

Framework applicable to vision, NLP, and reinforcement learning models

Abstract

Deep neural networks are able to solve tasks across a variety of domains and modalities of data. Despite many empirical successes, we lack the ability to clearly understand and interpret the learned internal mechanisms that contribute to such effective behaviors or, more critically, failure modes. In this work, we present a general method for visualizing an arbitrary neural network's inner mechanisms and their power and limitations. Our dataset-centric method produces visualizations of how a trained network attends to components of its inputs. The computed "attention masks" support improved interpretability by highlighting which input attributes are critical in determining output. We demonstrate the effectiveness of our framework on a variety of deep neural network architectures in domains from computer vision, natural language processing, and reinforcement learning. The primary contribution of our approach is an interpretable visualization of attention that provides unique insights into the network's underlying decision-making process irrespective of the data modality.