OpenLM Research·released 2023-06-07
Scored on 6 benchmarks. 0% of its scores are independently reproduced — the rest are self-reported or unverified.
Every score, ranked, colored by evidence status.
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| HellaSwag | 62.4accuracy (%) | self-reported· optimized | 2023-06-07 |
| PIQA | 52accuracy (%) | self-reported· optimized | 2023-06-07 |
| Winogrande | 34accuracy (%) | self-reported· optimized | 2023-06-07 |
| OpenBookQA | 18.67accuracy (%) | self-reported· optimized | 2023-06-07 |
| ARC AI2 | 18.27accuracy (%) | self-reported· optimized | 2023-06-07 |
| MMLU | 6.53accuracy (%) | self-reported· optimized | 2023-06-07 |