ColorShadow-7B-v3: benchmark results
A Mistral-family gradient SLERP merge by nlpguy. The release recipe specifies layer-dependent attention and feed-forward interpolation; the numbered versions use distinct recipes. Provider: nlpguy. Access: Open.
Aggregate Rating 1417 ± 37, rank #5318 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 62.88 | MC2 (%) (0-shot) | 79.8 |
| Open LLM Leaderboard v1 - ARC Challenge | 67.58 | Normalized accuracy (%) (25-shot) | 77.4 |
| Open LLM Leaderboard v1 - WinoGrande | 80.11 | Accuracy (%) (5-shot) | 72.4 |
| Open LLM Leaderboard v1 - HellaSwag | 85.04 | Normalized accuracy (%) (10-shot) | 71.3 |
| Open LLM Leaderboard v1 - GSM8K | 47.54 | Accuracy (%) (5-shot) | 60.8 |
| Open LLM Leaderboard v1 - MMLU | 60.57 | Accuracy (%) (5-shot) | 48.9 |
Interactive version: theaggregate.ai/model?slug=colorshadow-7b-v3 · How It Works · Data refreshed daily, snapshot 2026-11-08.