Of Non-Linearity and Commutativity in BERT
This provides insights into transformer design for NLP researchers, but is incremental as it builds on existing understanding of BERT.
The study analyzed the transformer architecture, particularly BERT, revealing that feed-forward networks (FFNs) are inefficient but crucial, and cannot be replaced by attention blocks without performance loss, while also showing BERT's inductive bias towards layer commutativity due to skip connections.
In this work we provide new insights into the transformer architecture, and in particular, its best-known variant, BERT. First, we propose a method to measure the degree of non-linearity of different elements of transformers. Next, we focus our investigation on the feed-forward networks (FFN) inside transformers, which contain 2/3 of the model parameters and have so far not received much attention. We find that FFNs are an inefficient yet important architectural element and that they cannot simply be replaced by attention blocks without a degradation in performance. Moreover, we study the interactions between layers in BERT and show that, while the layers exhibit some hierarchical structure, they extract features in a fuzzy manner. Our results suggest that BERT has an inductive bias towards layer commutativity, which we find is mainly due to the skip connections. This provides a justification for the strong performance of recurrent and weight-shared transformer models.