Multiple-choice reading comprehension on ViMMRC 2.0
60.32AccuracyViMultiChoice
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ViMultiChoiceExplanation Generator=w/o2026.02 | 60.32 | 60.23 | |
| ViCLSR2026.03 | 59.06 | — | |
| viBERTMulti-step Attention Network=true, Natural Language Inference=true2026.02 | 58.81 | — | |
| CafeBERT2026.03 | 57.98 | — | |
| XLM-Rvariant=Large2026.03 | 57.52 | — | |
| viBERTMulti-step Attention Network=true2026.02 | 57.17 | — | |
| CTM2026.02 | 57.17 | 57.17 | |
| mBERTNatural Language Inference=true2026.02 | 56.99 | — | |
| viBERTNatural Language Inference=true2026.02 | 56.81 | — | |
| ViT5 Encoder2026.02 | 56.63 | 56.54 | |
| OCN2026.02 | 56.27 | 56.29 | |
| DCMN+2026.02 | 55.73 | 55.71 | |
| mBERTMulti-step Attention Network=true, Natural Language Inference=true2026.02 | 55.64 | — | |
| PhoBERTvariant=Large2026.03 | 54.73 | — | |
| PhoBERTvariant=Base2026.03 | 53.92 | — | |
| viBERT2026.02 | 53.74 | — | |
| mBERT2026.03 | 53.38 | — | |
| mBERTMulti-step Attention Network=true2026.02 | 52.93 | — | |
| MMN2026.02 | 52.93 | 52.73 | |
| XLM-R2026.02 | 51.84 | — | |
| mBERT2026.02 | 47.79 | — | |
| DynSAN2026.02 | 44.36 | 43.74 | |
| BERT4News2026.02 | 41.74 | — | |
| CoMatch2026.02 | 41.21 | 40.55 | |
| DiffCSE2026.03 | 41.11 | — | |
| ElimiNET2026.02 | 38.59 | 38.21 | |
| CNN for MCQA2026.02 | 35.08 | 34.73 | |
| HAF2026.02 | 31.65 | 30.68 | |
| XLM-Rvariant=Base2026.03 | 29.39 | — |