Sina-Thor-7B-Merge: benchmark results
Azazelle's DARE-TIES merge, whose published recipe uses mistralai/Mistral-7B-v0.1 as its base and rishiraj/smol-7b, SanjiWatsuki/openchat-3.5-1210-starling-slerp, Azazelle/Dumb-Maidlet as merge inputs. Its published Mistral configuration specifies 32 decoder layers. Provider: Azazelle. Access: Open.
Aggregate Rating 1434 ± 34, rank #4504 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 68.69 | Accuracy (%) (5-shot) | 86.7 |
| Open LLM Leaderboard v1 - MMLU | 65.17 | Accuracy (%) (5-shot) | 82.5 |
| Open LLM Leaderboard v1 - HellaSwag | 85.69 | Normalized accuracy (%) (10-shot) | 76.1 |
| Open LLM Leaderboard v1 - WinoGrande | 80.51 | Accuracy (%) (5-shot) | 74.3 |
| Open LLM Leaderboard v1 - ARC Challenge | 66.21 | Normalized accuracy (%) (25-shot) | 72.7 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 50.01 | MC2 (%) (0-shot) | 46.8 |
Interactive version: theaggregate.ai/model?slug=sina-thor-7b-merge · How It Works · Data refreshed daily, snapshot 2026-11-08.