Open-domain Question Answering on SQUAD Open (test)
56.6Exact MatchIDRQA
Evaluation results
| Method | Links | ||
|---|---|---|---|
| IDRQA2020.09 | 56.6 | — | |
| Recurrent Graph-based Retrieval2019.11 | 56.5 | 63.8 | |
| Recurrent Retriever2020.09 | 56.5 | — | |
| multi-passage2019.11 | 53 | 60.9 | |
| Multi-passage BERT2020.09 | 53 | — | |
| Weaver2019.11 | 42.3 | — | |
| RE³2019.11 | 41.9 | 50.2 | |
| MUPPETgranularity=sentence-level2019.06 | 39.3 | 46.2 | |
| MUPPET2019.11 | 39.3 | 46.2 | |
| MUPPET2020.09 | 39.3 | — | |
| BERTserini2019.06 | 38.6 | 46.1 | |
| BERTserini2019.11 | 38.6 | 46.1 | |
| BERTserini2020.09 | 38.6 | — | |
| Dense Passage Retriever2020.09 | 36.7 | — | |
| DENSPI-hybrid2019.11 | 36.2 | 44.4 | |
| MUPPETgranularity=paragraph-level2019.06 | 35.6 | 42.5 | |
| Minimal2019.06 | 34.7 | 42.6 | |
| MINIMAL2019.11 | 34.7 | 42.5 | |
| TF-IDF + Reader2019.06 | 34.6 | 41.6 | |
| BM25+BERT2019.11 | 33.2 | — | |
| Multi-step2019.06 | 31.9 | 39.2 | |
| Multi-step Reasoner2019.11 | 31.9 | 39.2 | |
| Multi-step Reasoner2020.09 | 31.9 | — | |
| Par. Ranker + Full Agg.2019.06 | 30.2 | — | |
| Paragraph Ranker2019.11 | 30.2 | — | |
| DrQAmode=multitask2019.06 | 29.8 | — | |
| DrQA2020.09 | 29.8 | — | |
| DrQA2019.11 | 29.3 | — | |
| Reinforced Ranker-Reader2017.08 | 29.1 | 37.5 | |
| R^32019.06 | 29.1 | 37.5 | |
| R³2019.11 | 29.1 | 37.5 | |
| R32020.09 | 29.1 | — | |
| DS-QA2019.06 | 28.7 | 36.6 | |
| DrQA2017.08 | 28.4 | — | |
| DrQA2019.06 | 28.4 | — | |
| Simple Ranker-Reader2017.08 | 27.2 | 35.8 | |
| Single Reader2017.08 | 26.9 | 35.4 | |
| ORQA2019.11 | 20.2 | — | |
| ORQA2020.09 | 20.2 | — | |
| DrQA-MTLAdditional Resources=multiple training datasets2017.08 | — | 29.8 |