Answer Sentence Selection on ASNQ
70.5P@1ELECTRA-Base + SSP (DPC)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| ELECTRA-Base + SSP (DPC)Context=true2023.05 | 70.5 | 74.4 | — | |
| ELECTRA-Base + SSP (DSLC)Context=true2023.05 | 70 | 74.3 | — | |
| ELECTRA-Base + SSP (SDC)Context=true2023.05 | 69.6 | 74.7 | — | |
| ROBERTa-Base + SSP (DPC)Context=true2023.05 | 69.2 | 73.2 | — | |
| ROBERTa-Base + SSP (DSLC)Context=true2023.05 | 69 | 72.9 | — | |
| ELECTRA-Base + SSP (All)Context=true2023.05 | 68.8 | 73.8 | — | |
| ROBERTa-Base + SSP (SDC)Context=true2023.05 | 68.7 | 73.1 | — | |
| ROBERTa-Base + SSP (All)Context=true2023.05 | 68.2 | 72.9 | — | |
| ELECTRA-BaseContext=true2023.05 | 68.1 | 72.3 | — | |
| ROBERTa-BaseContext=true2023.05 | 67.6 | 71.6 | — | |
| ROBERTaBase Flexible (REk) + MSPPBackbone=RoBERTa-base, Architecture Type=Flexible, Prediction Head=REk, Pre-training Objective=MSPP2023.09 | 65.3 | 69.4 | 75.7 | |
| ELECTRA-BaseContext=false2023.05 | 65 | 69.3 | — | |
| Joint MSPP AEkArchitecture=Joint model, Pre-training Task=MSPP, Pre-training Head=AEk, Fine-tuning Head=AEk, Protocol=Re-ranking top-52022.05 | 64.3 | — | — | |
| Joint MSPP IEkArchitecture=Joint model, Pre-training Task=MSPP, Pre-training Head=IEk, Fine-tuning Head=IEk, Protocol=Re-ranking top-52022.05 | 63.9 | — | — | |
| ROBERTa-BaseContext=false2023.05 | 63.5 | 68.2 | — | |
| ROBERTaBase Fixed (IEk) + MSPPBackbone=RoBERTa-base, Architecture Type=Fixed, Prediction Head=IEk, Pre-training Objective=MSPP2023.09 | 63 | 67.2 | 73.7 | |
| ROBERTaBase Fixed (AEk) + MSPPBackbone=RoBERTa-base, Architecture Type=Fixed, Prediction Head=AEk, Pre-training Objective=MSPP2023.09 | 63 | 67.3 | 73.7 | |
| Pairwise RoBERTa-BaseArchitecture=Pairwise Cross-encoder, Protocol=Re-ranking top-52022.05 | 61.8 | — | — | |
| ROBERTaBaseBackbone=RoBERTa-base, Reference=Liu et al., 2019b2023.09 | 61.8 | 66.9 | 73.1 | |
| ROBERTaBase Flexible (REk)Backbone=RoBERTa-base, Architecture Type=Flexible, Prediction Head=REk, Pre-training Objective=Standard2023.09 | 58.9 | 63.7 | 70.2 | |
| ROBERTaBase Fixed (AEk)Backbone=RoBERTa-base, Architecture Type=Fixed, Prediction Head=AEk, Pre-training Objective=Standard2023.09 | 57.7 | 62.3 | 66 | |
| ROBERTaBase Fixed (IEk)Backbone=RoBERTa-base, Architecture Type=Fixed, Prediction Head=IEk, Pre-training Objective=Standard2023.09 | 57.4 | 62.1 | 65.6 | |
| Joint RoBERTa-BaseArchitecture=Joint model, Fine-tuning Head=AEk, Protocol=Re-ranking top-52022.05 | 25.4 | — | — | |
| Joint RoBERTa-BaseArchitecture=Joint model, Fine-tuning Head=IEk, Protocol=Re-ranking top-52022.05 | 25.2 | — | — |