Skip to main content

Vocabulary of natural language processing

Search from vocabulary

  • Help
align
annotate
download

Concept information

... > NLP methods and tools > model > statistical model > foundation model > language model > PIXEL
... > NLP methods and tools > statistics > statistical model > foundation model > language model > PIXEL

Preferred term

PIXEL  

Definition

  • 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. (Rust et al., 2022)

Broader concept

Synonym(s)

  • picture element
  • pixel-based encoders of language
  • px

Example

  • Note that PIXEL is a 112M encoder-decoder model but only the 86M encoder is used in this experiment and GPT-2 uses even more parameters (126M). (Tai, Liao, Suglia & Vergari, 2024)
  • PIXEL transforms its input text into RGB images and is a pre-trained vision transformer masked autoencoder with 112M parameters. (Fröbe, Stein, Gollub, Hagen & Potthast, 2023)

In other languages

  • PIXEL

    French

  • pixel-based encoders of language
  • px

URI

http://data.loterre.fr/ark:/67375/8LP-WS2JC6KS-7

Download this concept:

RDF/XML TURTLE JSON-LD Last modified 4/26/24