Sentence Classification on MPQA (test)
90.8AccuracyEFL
Evaluation results
| Method | Links | |
|---|---|---|
| EFLTraining setup=Full training dataset, Backbone=RoBERTa-large2021.04 | 90.8 | |
| AdaSentevaluation=10-fold cross validation2017.09 | 90.4 | |
| DiSANevaluation=10-fold cross validation2017.09 | 90.1 | |
| SRUevaluation=10-fold cross validation2017.09 | 89.7 | |
| Fine-tuningTraining setup=Full training dataset, Backbone=RoBERTa-large2021.04 | 89.4 | |
| Wide CNNsevaluation=10-fold cross validation2017.09 | 88.8 | |
| Skip-thoughtevaluation=10-fold cross validation2017.09 | 87.5 | |
| EFLTraining setup=Few-shot, K (samples per class)=8, Backbone=RoBERTa-large2021.04 | 87 | |
| cBoWevaluation=10-fold cross validation2017.09 | 86.4 | |
| KNN-Promptingprivacy budget (ε)=Non-private, number of hyperplanes=24, number of hash tables=42026.05 | 85.4 | |
| DP SimHash Bucketingprivacy budget (ε)=1, number of hyperplanes=24, number of hash tables=42026.05 | 84.5 | |
| DP SimHash Bucketingprivacy budget (ε)=3, number of hyperplanes=24, number of hash tables=42026.05 | 84.5 | |
| DP SimHash Bucketingprivacy budget (ε)=7, number of hyperplanes=24, number of hash tables=42026.05 | 84.5 | |
| DP SimHash Bucketingprivacy budget (ε)=8, number of hyperplanes=24, number of hash tables=42026.05 | 84.5 | |
| DP SimHash Bucketingprivacy budget (ε)=5, number of hyperplanes=24, number of hash tables=42026.05 | 84.4 | |
| DP SimHash Bucketingprivacy budget (ε)=0.1, number of hyperplanes=24, number of hash tables=42026.05 | 83.7 | |
| Stilts-CloseTraining setup=Few-shot, K (samples per class)=8, Backbone=RoBERTa-large2021.04 | 76.6 | |
| LM-BFFTraining setup=Few-shot, K (samples per class)=8, Backbone=RoBERTa-large2021.04 | 74.2 | |
| Stilts-NLITraining setup=Few-shot, K (samples per class)=8, Backbone=RoBERTa-large2021.04 | 67.7 | |
| EFL wo PTTraining setup=Few-shot, K (samples per class)=8, Backbone=RoBERTa-large2021.04 | 60.4 | |
| Fine-tuningTraining setup=Few-shot, K (samples per class)=8, Backbone=RoBERTa-large2021.04 | 59 | |
| MajorityTraining setup=Full training dataset2021.04 | 50 |