phi-2-vllm: benchmark results
Microsoft's Phi-2 base transformer, trained on synthetic NLP text and filtered websites without RLHF fine-tuning. This row preserves its separately evaluated revision, precision, shot count or runtime configuration. Provider: Microsoft. Access: Open.
Aggregate Rating 1365 ± 40, rank #7303 of 10916 rated models, from 11 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| URIAL-Bench - Coding | 4.55 | Judge Score (0-10) | 72.2 |
| URIAL-Bench - Reasoning | 5.5 | Judge Score (0-10) | 72.2 |
| URIAL-Bench - Overall | 6.06 | Judge Score (0-10) | 61.1 |
| URIAL-Bench - Turn 1 | 7.17 | Judge Score (0-10) | 61.1 |
| URIAL-Bench - Extraction | 5.3 | Judge Score (0-10) | 55.6 |
| URIAL-Bench - Turn 2 | 4.94 | Judge Score (0-10) | 55.6 |
| URIAL-Bench - Writing | 7.45 | Judge Score (0-10) | 55.6 |
| URIAL-Bench - Humanities | 8.65 | Judge Score (0-10) | 50 |
| URIAL-Bench - Math | 3.35 | Judge Score (0-10) | 50 |
| URIAL-Bench - STEM | 7.11 | Judge Score (0-10) | 50 |
| URIAL-Bench - Roleplay | 6.63 | Judge Score (0-10) | 41.7 |
Interactive version: theaggregate.ai/model?slug=phi-2-vllm · How It Works · Data refreshed daily, snapshot 2026-11-08.