GPT Neox 20B — Benchmarks
Benchmark scores for GPT Neox 20B aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.
Knowledge
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| HellaSwag | 70.5% | #32 / 42 | #55 / 76 |
| ARC Challenge | 41.1% | #39 / 51 | #60 / 77 |
| BoolQ | 64.9% | #28 / 33 | #64 / 77 |
| OpenBookQA | 38.8% | #16 / 21 | #34 / 42 |
Reasoning
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| WinoGrande | 66.1% | #29 / 46 | #58 / 80 |
| PIQA | 76.7% | #27 / 35 | #47 / 60 |
Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.