Mistral-7B-Instruct-v0.2-attention-sparsity-10-v0.1: benchmark results
Provider: Mistral.
Aggregate Rating 1401 ± 34, rank #6049 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 68.11 | MC2 (%) (0-shot) | 87 |
| Open LLM Leaderboard v1 - HellaSwag | 84.88 | Normalized accuracy (%) (10-shot) | 69.7 |
| Open LLM Leaderboard v1 - ARC Challenge | 63.05 | Normalized accuracy (%) (25-shot) | 61 |
| Open LLM Leaderboard v1 - GSM8K | 38.89 | Accuracy (%) (5-shot) | 53.7 |
| Open LLM Leaderboard v1 - WinoGrande | 77.11 | Accuracy (%) (5-shot) | 53.3 |
| Open LLM Leaderboard v1 - MMLU | 60.84 | Accuracy (%) (5-shot) | 49.9 |
Interactive version: theaggregate.ai/model?slug=mistral-7b-instruct-v0-2-attention-sparsity-10-v0-1 · How It Works · Data refreshed daily, snapshot 2026-11-08.