Text Classification on SST binary
91.7Accuracybyte mLSTM
Evaluation results
| Method | Links | |
|---|---|---|
| byte mLSTMd*=40962018.05 | 91.7 | |
| SBERT2025.12 | 90.7 | |
| CDME*-SigmoidNumber of parameters=4.6M, embedding sets=multiple2018.04 | 89.8 | |
| NSE2018.04 | 89.7 | |
| DCG2018.04 | 89.4 | |
| CDME*-SoftmaxNumber of parameters=4.6M, embedding sets=multiple2018.04 | 89.3 | |
| CDMENumber of parameters=4.1M2018.04 | 89.2 | |
| Unweighted DMENumber of parameters=4.1M2018.04 | 89 | |
| DMENumber of parameters=4.1M2018.04 | 88.7 | |
| DMN2018.04 | 88.6 | |
| Naive baselineNumber of parameters=5.4M2018.04 | 88.5 | |
| Const. Tree LSTM2018.04 | 88 | |
| GloVe BiLSTM-MaxNumber of parameters=4.1M2018.04 | 88 | |
| MC-QTd*=48002018.05 | 87.6 | |
| à la carten=3, d*=48002018.05 | 86.7 | |
| FastText BiLSTM-MaxNumber of parameters=4.1M2018.04 | 86.7 | |
| à la carten=2, d*=32002018.05 | 85.8 | |
| DisCn=2-3, d*=3200-48002018.05 | 85.5 | |
| skip-thoughtsd*=48002018.05 | 85.1 | |
| DPCLie2025.12 | 85.1 | |
| SCLie2025.12 | 84.3 | |
| à la carten=1, d*=16002018.05 | 84.1 | |
| SCNN2025.12 | 83.5 | |
| LinearArchitecture=Fully connected Neural Network replacement for Convolution layer2025.12 | 83.1 | |
| DPCNN2025.12 | 82.1 | |
| BonGn=2, d*=V1 + V22018.05 | 80.9 | |
| BonGn=1, d*=V12018.05 | 80.7 | |
| Sent2Vecn=1-2, d*=7002018.05 | 80.2 | |
| BonGn=3, d*=V1 + V2 + V32018.05 | 80.1 |