Visualizing attention zones in machine reading comprehension models
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914993151148032 |
|---|---|
| author | Cui, Yiming Zhang, Wei-Nan Liu, Ting |
| author_facet | Cui, Yiming Zhang, Wei-Nan Liu, Ting |
| contents | The attention mechanism plays an important role in the machine reading comprehension (MRC) model. Here, we describe a pipeline for building an MRC model with a pretrained language model and visualizing the effect of each attention zone in different layers, which can indicate the explainability of the model. With the presented protocol and accompanying code, researchers can easily visualize the relevance of each attention zone in the MRC model. This approach can be generalized to other pretrained language models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_20652 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Visualizing attention zones in machine reading comprehension models Cui, Yiming Zhang, Wei-Nan Liu, Ting Computation and Language The attention mechanism plays an important role in the machine reading comprehension (MRC) model. Here, we describe a pipeline for building an MRC model with a pretrained language model and visualizing the effect of each attention zone in different layers, which can indicate the explainability of the model. With the presented protocol and accompanying code, researchers can easily visualize the relevance of each attention zone in the MRC model. This approach can be generalized to other pretrained language models. |
| title | Visualizing attention zones in machine reading comprehension models |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2410.20652 |