| Benchmark | Model | Rank | Results |
|---|---|---|---|
| image-classification-on-cifar-10 | PREBR++JS@70% | – | Percentage correct: 98.8 |
| image-classification-on-cifar-100 | PREBR++JS@70% | – | Percentage correct: 92.4 |
| natural-language-inference-on-rte-glue-dev | PREBR@80% (batch size 8) | – | Accuracy: 72.92 |