Language Modelling with Pixels

Phillip Rust, Jonas F. Lotz, Emanuele Bugliarello, Elizabeth Salesky, Miryam de Lhoneux, Desmond Elliott

Publikation: Bidrag til bog/antologi/rapportKonferencebidrag i proceedingsForskningpeer review

18 Downloads (Pure)

Abstract

Language models are defined over a finite set of inputs, which creates a vocabulary bottleneck when we attempt to scale the number of supported languages. Tackling this bottleneck results in a trade-off between what can be represented in the embedding matrix and computational issues in the output layer. This paper introduces PIXEL, the Pixel-based Encoder of Language, which suffers from neither of these issues. PIXEL is a pretrained language model that renders text as images, making it possible to transfer representations across languages based on orthographic similarity or the co-activation of pixels. PIXEL is trained to reconstruct the pixels of masked patches instead of predicting a distribution over tokens. We pretrain the 86M parameter PIXEL model on the same English data as BERT and evaluate on syntactic and semantic tasks in typologically diverse languages, including various non-Latin scripts. We find that PIXEL substantially outperforms BERT on syntactic and semantic processing tasks on scripts that are not found in the pretraining data, but PIXEL is slightly weaker than BERT when working with Latin scripts. Furthermore, we find that PIXEL is more robust than BERT to orthographic attacks and linguistic code-switching, further confirming the benefits of modelling language with pixels.
OriginalsprogEngelsk
TitelThe Eleventh International Conference on Learning Representations
Antal sider32
ForlagarXiv.org
Publikationsdato2023
StatusUdgivet - 2023
Begivenhed11h International Conference on Learning Representations - ICLR 2023 - Kigali, Rwanda
Varighed: 1 maj 20235 maj 2023

Konference

Konference11h International Conference on Learning Representations - ICLR 2023
Land/OmrådeRwanda
ByKigali
Periode01/05/202305/05/2023

Citationsformater