Dropout during inference as a model for neurological degeneration in an image captioning network

TL;DR

This paper explores using dropout during inference in an image captioning model to simulate neurological degeneration effects, revealing that moderate dropout best mimics disease-related language changes.

Contribution

It introduces a novel method of simulating neurodegenerative effects in neural networks by applying dropout during inference, providing insights into language deterioration.

Findings

Moderate dropout (0.4) during inference best replicates training corpus word distribution.

Dropout during inference affects linguistic metrics and language production.

Simulation of neurodegeneration effects in neural networks using dropout.

Abstract

We replicate a variation of the image captioning architecture by Vinyals et al. (2015), then introduce dropout during inference mode to simulate the effects of neurodegenerative diseases like Alzheimer's disease (AD) and Wernicke's aphasia (WA). We evaluate the effects of dropout on language production by measuring the KL-divergence of word frequency distributions and other linguistic metrics as dropout is added. We find that the generated sentences most closely approximate the word frequency distribution of the training corpus when using a moderate dropout of 0.4 during inference.