Impact of pre-training on background knowledge and societal bias
<p>With appropriate pre-training on unstructured text, larger and more accurate neural network models can be trained. Unfortunately, unstructured pre-training data may contain undesired societal biases, which a model may mimic and amplify. This thesis focuses on both improving unsuperv...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2021
|
| Subjects: |