Blurstral-7B-slerp: benchmark results
liminerity's Blurstral-7b-slerp is a SLERP merge of mistralai/Mistral-7B-v0.1, liminerity/Blur-7b-slerp-v0.1. The released configuration uses 32 mistral decoder layers; merge recipe dtype is not a measured evaluator precision. Provider: liminerity. Access: Open.
Aggregate Rating 1434 ± 33, rank #4445 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 65.18 | Accuracy (%) (5-shot) | 82.7 |
| Open LLM Leaderboard v1 - WinoGrande | 81.37 | Accuracy (%) (5-shot) | 78.9 |
| Open LLM Leaderboard v1 - GSM8K | 62.85 | Accuracy (%) (5-shot) | 76.2 |
| Open LLM Leaderboard v1 - HellaSwag | 85.38 | Normalized accuracy (%) (10-shot) | 74.1 |
| Open LLM Leaderboard v1 - ARC Challenge | 66.3 | Normalized accuracy (%) (25-shot) | 73 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 53.4 | MC2 (%) (0-shot) | 59.5 |
Interactive version: theaggregate.ai/model?slug=blurstral-7b-slerp · How It Works · Data refreshed daily, snapshot 2026-11-08.