| Benchmark | Model | Rank | Results |
|---|---|---|---|
| common-sense-reasoning-on-winogrande | Titans + GRM | – | Accuracy: 57.4 |
| language-modelling-on-lambada | Titans + GRM | – | Perplexity: 11.29 |
| question-answering-on-boolq | Titans + GRM | – | Accuracy: 43.9 |
| sentence-completion-on-hellaswag | Titans + GRM | – | Accuracy: 74.5 |
| social-commonsense-reasoning-on-siqa | Titans + GRM | – | Accuracy: 42.6 |