Cargando…
Learning Numerosity Representations with Transformers: Number Generation Tasks and Out-of-Distribution Generalization
One of the most rapidly advancing areas of deep learning research aims at creating models that learn to disentangle the latent factors of variation from a data distribution. However, modeling joint probability mass functions is usually prohibitive, which motivates the use of conditional models assum...
| Autores principales: | , , |
|---|---|
| Formato: | Online Artículo Texto |
| Lenguaje: | English |
| Publicado: |
MDPI
2021
|
| Materias: | |
| Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8303966/ https://www.ncbi.nlm.nih.gov/pubmed/34356398 http://dx.doi.org/10.3390/e23070857 |