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.
Original language | English |
---|---|
Publication status | Published - 2019 |
Event | 7th International Conference on Learning Representations, ICLR 2019 - New Orleans, United States Duration: 6 May 2019 → 9 May 2019 |
Conference
Conference | 7th International Conference on Learning Representations, ICLR 2019 |
---|---|
Country/Territory | United States |
City | New Orleans |
Period | 6/05/19 → 9/05/19 |