Cascaded Head-colliding Attention

05/31/2021 ∙ by Lin Zheng, et al. ∙ 0

Transformers have advanced the field of natural language processing (NLP) on a variety of important tasks. At the cornerstone of the Transformer architecture is the multi-head attention (MHA) mechanism which models pairwise interactions between the elements of the sequence. Despite its massive success, the current framework ignores interactions among different heads, leading to the problem that many of the heads are redundant in practice, which greatly wastes the capacity of the model. To improve parameter efficiency, we re-formulate the MHA as a latent variable model from a probabilistic perspective. We present cascaded head-colliding attention (CODA) which explicitly models the interactions between attention heads through a hierarchical variational distribution. We conduct extensive experiments and demonstrate that CODA outperforms the transformer baseline, by 0.6 perplexity on in language modeling, and by 0.6 BLEU on in machine translation, due to its improvements on the parameter efficiency.[Our implementation is publicly available at <>.]



There are no comments yet.


page 7

page 13

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.