LlamaReflect-8B-CoT: benchmark results
Provider: Meta. Access: Open.
Aggregate Rating 1403 ± 34, rank #5941 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 65.38 | Accuracy (%) (5-shot) | 84.2 |
| Open LLM Leaderboard v1 - GSM8K | 44.12 | Accuracy (%) (5-shot) | 58.2 |
| Open LLM Leaderboard v1 - WinoGrande | 77.58 | Accuracy (%) (5-shot) | 57 |
| Open LLM Leaderboard v1 - HellaSwag | 81.82 | Normalized accuracy (%) (10-shot) | 47.2 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 47.84 | MC2 (%) (0-shot) | 38.8 |
| Open LLM Leaderboard v1 - ARC Challenge | 57.17 | Normalized accuracy (%) (25-shot) | 36.8 |
Interactive version: theaggregate.ai/model?slug=llamareflect-8b-cot · How It Works · Data refreshed daily, snapshot 2026-11-08.