Identifying and Controlling Important Neurons in Neural Machine Translation

TL;DR

This paper introduces unsupervised methods to identify and manipulate key neurons in neural machine translation models, revealing their role in linguistic representation and enabling controllable translation outputs.

Contribution

It presents novel unsupervised techniques for discovering important neurons in NMT models and demonstrates how to control translations by modifying neuron activations.

Findings

Translation quality depends on identified neurons.

Many neurons capture linguistic phenomena.

Controlling neuron activations influences translation outputs.

Abstract

Neural machine translation (NMT) models learn representations containing substantial linguistic information. However, it is not clear if such information is fully distributed or if some of it can be attributed to individual neurons. We develop unsupervised methods for discovering important neurons in NMT models. Our methods rely on the intuition that different models learn similar properties, and do not require any costly external supervision. We show experimentally that translation quality depends on the discovered neurons, and find that many of them capture common linguistic phenomena. Finally, we show how to control NMT translations in predictable ways, by modifying activations of individual neurons.