Cloze Prompt Task on mLAMA (test)
42.4Accuracy (ar)mLUKE-Ebase ([X] & [Y])
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| mLUKE-Ebase ([X] & [Y])representation_type=entity-based, entity_prediction=candidate entity in [Y], additional_input=entity token of [X]2021.10 | 42.4 | 47.5 | 44.2 | 35.9 | 56.2 | 40.3 | 35.5 | 55.2 | 46.7 | 44.9 | |
| mLUKE-Ebase ([Y])representation_type=entity-based, entity_prediction=candidate entity in [Y]2021.10 | 27.8 | 37.5 | 30.4 | 28.4 | 44.2 | 28.9 | 25.8 | 42.1 | 33.4 | 33.2 | |
| mLUKE-Wbaserepresentation_type=word-based2021.10 | 22.3 | 31.3 | 18.4 | 19.6 | 46.7 | 18.4 | 16.7 | 31.9 | 29.3 | 26.1 | |
| XLM-Rbase + extra trainingtraining_corpus=CommonCrawl + Wikipedia2021.10 | 21.2 | 35 | 23 | 22.2 | 46.8 | 19.6 | 17.5 | 34.4 | 30.7 | 27.8 | |
| mBERTtype=word-based baseline2021.10 | 17.1 | 36.8 | 24 | 24.3 | 42.9 | 14.3 | 19.5 | 39.4 | 26.2 | 27.2 | |
| XLM-Rbasetraining_corpus=CommonCrawl2021.10 | 14.2 | 27.2 | 16.2 | 14.9 | 28.2 | 11.9 | 11.7 | 25.1 | 17.6 | 18.5 |