llama-2-alpacagpt4-1000step: benchmark results
Provider: Meta.
Aggregate Rating 1386 ± 19, rank #6670 of 10916 rated models, from 62 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MMLU-by-task - Global Facts | 49 | Accuracy (%) | 97.7 |
| MMLU-by-task - Machine Learning | 52.68 | Accuracy (%) | 97.1 |
| MMLU-by-task - Virology | 54.22 | Accuracy (%) | 96.7 |
| MMLU-by-task - TruthfulQA MC1 | 40.39 | Accuracy (%) | 96 |
| MMLU-by-task - US Foreign Policy | 89 | Accuracy (%) | 95.3 |
| MMLU-by-task - TruthfulQA MC2 | 55.57 | Accuracy (%) | 93.7 |
| MMLU-by-task - Human Aging | 74.44 | Accuracy (%) | 93.5 |
| MMLU-by-task - Philosophy | 74.6 | Accuracy (%) | 93.5 |
| MMLU-by-task - Conceptual Physics | 61.7 | Accuracy (%) | 93.4 |
| MMLU-by-task - Business Ethics | 67 | Accuracy (%) | 92.9 |
| MMLU-by-task - High School Geography | 83.33 | Accuracy (%) | 92.9 |
| MMLU-by-task - Professional Psychology | 69.12 | Accuracy (%) | 92.9 |
Interactive version: theaggregate.ai/model?slug=llama-2-alpacagpt4-1000step · How It Works · Data refreshed daily, snapshot 2026-11-08.