Improving model adversarial robustness in Extractive Question Answering via Wasserstein-Guided feature Representations
Extractive Question Answering (EQA) models aim to locate accurate answers from passages given a question but are highly susceptible to adversarial attacks. Existing adversarial training methods improve robustness by generating perturbed passages, yet they remain computationally expensive and prone t...
Na minha lista:
| Principais autores: | , , |
|---|---|
| Formato: | Artigo |
| Idioma: | Inglês |
| Publicado em: |
Elsevier
2025-09-01
|
| coleção: | Alexandria Engineering Journal |
| Assuntos: | |
| Acesso em linha: | http://www.sciencedirect.com/science/article/pii/S1110016825006155 |
| Tags: |
Sem tags, seja o primeiro a adicionar uma tag!
|