BoolQ, ARC, WinoGrande, HellaSwag
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Zero-shot Language Understanding | BoolQ, ARC-e, ARC-c, WinoGrande, HellaSwag | ARC-e Accuracy83.08 | 8 | |
| Natural Language Reasoning | BoolQ, ARC-e, ARC-c, WinoGrande (WinoG), HellaSwag (HelloS) | BoolQ Accuracy75.2 | 4 |
Showing 2 of 2 rows