Papers
arxiv:2103.01075

OmniNet: Omnidirectional Representations from Transformers

Published on Mar 1, 2021
Authors:
,
,
,
,
,
,
,
,

Abstract

Omnidirectional Representations from Transformers (OmniNet) enhance performance across various tasks by allowing tokens to attend to all tokens in the network, using efficient self-attention models as a meta-learner.

AI-generated summary

This paper proposes Omnidirectional Representations from Transformers (OmniNet). In OmniNet, instead of maintaining a strictly horizontal receptive field, each token is allowed to attend to all tokens in the entire network. This process can also be interpreted as a form of extreme or intensive attention mechanism that has the receptive field of the entire width and depth of the network. To this end, the omnidirectional attention is learned via a meta-learner, which is essentially another self-attention based model. In order to mitigate the computationally expensive costs of full receptive field attention, we leverage efficient self-attention models such as kernel-based (Choromanski et al.), low-rank attention (Wang et al.) and/or Big Bird (Zaheer et al.) as the meta-learner. Extensive experiments are conducted on autoregressive language modeling (LM1B, C4), Machine Translation, Long Range Arena (LRA), and Image Recognition. The experiments show that OmniNet achieves considerable improvements across these tasks, including achieving state-of-the-art performance on LM1B, WMT'14 En-De/En-Fr, and Long Range Arena. Moreover, using omnidirectional representation in Vision Transformers leads to significant improvements on image recognition tasks on both few-shot learning and fine-tuning setups.

Community

Sign up or log in to comment

Models citing this paper 1

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2103.01075 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2103.01075 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.