M-Prometheus-14B: benchmark results
Provider: Other.
Aggregate Rating 1578 ± 41, rank #1283 of 10916 rated models, from 12 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Plan-RewardBench - Tool Irrelevance | 63.89 | Pairwise accuracy (%) on the 275 tool-irrelevance or tool-un | 65 |
| Plan-RewardBench - Safety Refusal | 62.44 | Pairwise accuracy (%) on the 51 safety-refusal pairs of the | 60 |
| IF-RewardBench - Overall Assessment (System-Prompt) | 0.05 | Kendall tau-b (-1 to 1) between the judge ranking of the res | 56.2 |
| Plan-RewardBench - Multi-Turn Planning (Hard) | 58.75 | Pairwise accuracy (%) on the 73 hard multi-turn planning pai | 30 |
| Plan-RewardBench - Single-Turn Planning (Easy) | 72.34 | Pairwise accuracy (%) on the 144 easy single-turn planning p | 30 |
| Plan-RewardBench - Single-Turn Planning (Hard) | 68.4 | Pairwise accuracy (%) on the 158 hard single-turn planning p | 30 |
| Plan-RewardBench | 65.08 | Pairwise accuracy (%), macro-average of the seven scenario c | 25 |
| Plan-RewardBench - Error Recovery | 64.55 | Pairwise accuracy (%) on the 361 robust error-recovery pairs | 20 |
| Plan-RewardBench - Multi-Turn Planning (Easy) | 65.23 | Pairwise accuracy (%) on the 109 easy multi-turn planning pa | 20 |
| IF-RewardBench - Overall Assessment | 0.02 | Kendall tau-b (-1 to 1) between the judge ranking of the res | 0 |
| IF-RewardBench - Overall Assessment (Multi-Turn) | 0.01 | Kendall tau-b (-1 to 1) between the judge ranking of the res | 0 |
| IF-RewardBench - Overall Assessment (Single-Turn) | 0.01 | Kendall tau-b (-1 to 1) between the judge ranking of the res | 0 |
Interactive version: theaggregate.ai/model?slug=m-prometheus-14b · How It Works · Data refreshed daily, snapshot 2026-11-08.