Llemma-PRM800K-7B: benchmark results
Provider: Other.
Aggregate Rating 1401 ± 45, rank #6072 of 10916 rated models, from 11 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| GR-Ben - Abductive Reasoning | 13.2 | F1 (%) of first-error identification: the harmonic mean of t | 38.1 |
| GR-Ben - Mixed-Form Reasoning | 20.1 | F1 (%) of first-error identification: the harmonic mean of t | 33.3 |
| GR-Ben - Deductive Reasoning | 13.1 | F1 (%) of first-error identification: the harmonic mean of t | 26.2 |
| PRMBench | 52 | Overall Score | 20 |
| GR-Ben - Analogical Reasoning | 18.8 | F1 (%) of first-error identification: the harmonic mean of t | 14.3 |
| GR-Ben - Inductive Reasoning | 17.1 | F1 (%) of first-error identification: the harmonic mean of t | 14.3 |
| GR-Ben | 16.1 | Average (self-reported) | 9.5 |
| GR-Ben - Chemistry | 21.5 | F1 (%) of first-error identification: the harmonic mean of t | 9.5 |
| GR-Ben - Computer Science | 13.2 | F1 (%) of first-error identification: the harmonic mean of t | 9.5 |
| GR-Ben - Physics | 17.3 | F1 (%) of first-error identification: the harmonic mean of t | 4.8 |
| GR-Ben - Biology | 10.4 | F1 (%) of first-error identification: the harmonic mean of t | 0 |
Interactive version: theaggregate.ai/model?slug=llemma-prm800k-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.