| Benchmark | Model | Rank | Results |
|---|---|---|---|
| multiple-choice-question-answering-mcqa-on-arc-c | LoopedLM-P2 learned-depth lambda_H=0.01 (1.6B) | #18 | Accuracy: 40.44 |
| language-modelling-on-lambada | LoopedLM-P2 learned-depth lambda_H=0.01 (1.6B) | #37 | Accuracy: 55.17 |
| question-answering-on-piqa | LoopedLM-P2 learned-depth lambda_H=0.01 (1.6B) | #115 | Accuracy: 73.72 |