EIT: Enhanced Interactive Transformer

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zheng, Tong, Li, Bei, Bao, Huiwen, Xiao, Tong, Zhu, Jingbo
Natura: Preprint
Pubblicazione: 2022
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866929373712482304
author Zheng, Tong
Li, Bei
Bao, Huiwen
Xiao, Tong
Zhu, Jingbo
author_facet Zheng, Tong
Li, Bei
Bao, Huiwen
Xiao, Tong
Zhu, Jingbo
contents Two principles: the complementary principle and the consensus principle are widely acknowledged in the literature of multi-view learning. However, the current design of multi-head self-attention, an instance of multi-view learning, prioritizes the complementarity while ignoring the consensus. To address this problem, we propose an enhanced multi-head self-attention (EMHA). First, to satisfy the complementary principle, EMHA removes the one-to-one mapping constraint among queries and keys in multiple subspaces and allows each query to attend to multiple keys. On top of that, we develop a method to fully encourage consensus among heads by introducing two interaction models, namely inner-subspace interaction and cross-subspace interaction. Extensive experiments on a wide range of language tasks (e.g., machine translation, abstractive summarization and grammar correction, language modeling), show its superiority, with a very modest increase in model size. Our code would be available at: https://github.com/zhengkid/EIT-Enhanced-Interactive-Transformer.
format Preprint
id arxiv_https___arxiv_org_abs_2212_10197
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle EIT: Enhanced Interactive Transformer
Zheng, Tong
Li, Bei
Bao, Huiwen
Xiao, Tong
Zhu, Jingbo
Computation and Language
Two principles: the complementary principle and the consensus principle are widely acknowledged in the literature of multi-view learning. However, the current design of multi-head self-attention, an instance of multi-view learning, prioritizes the complementarity while ignoring the consensus. To address this problem, we propose an enhanced multi-head self-attention (EMHA). First, to satisfy the complementary principle, EMHA removes the one-to-one mapping constraint among queries and keys in multiple subspaces and allows each query to attend to multiple keys. On top of that, we develop a method to fully encourage consensus among heads by introducing two interaction models, namely inner-subspace interaction and cross-subspace interaction. Extensive experiments on a wide range of language tasks (e.g., machine translation, abstractive summarization and grammar correction, language modeling), show its superiority, with a very modest increase in model size. Our code would be available at: https://github.com/zhengkid/EIT-Enhanced-Interactive-Transformer.
title EIT: Enhanced Interactive Transformer
topic Computation and Language
url https://arxiv.org/abs/2212.10197