NEBULA-XB-v1.0: benchmark results
A second-stage depth-upscaling release by TeeZee with 108 layers, derived through the SlimOrca-fine-tuned GALAXY v03 model. The recipe removes 18 layers from each source before combining them. Provider: TeeZee. Access: Open.
Aggregate Rating 1377 ± 36, rank #6939 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - WinoGrande | 77.66 | Accuracy (%) (5-shot) | 57.6 |
| Open LLM Leaderboard v1 - MMLU | 60.98 | Accuracy (%) (5-shot) | 50.4 |
| Open LLM Leaderboard v1 - HellaSwag | 81.78 | Normalized accuracy (%) (10-shot) | 47 |
| Open LLM Leaderboard v1 - ARC Challenge | 56.66 | Normalized accuracy (%) (25-shot) | 35.9 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 44.03 | MC2 (%) (0-shot) | 25.7 |
| Open LLM Leaderboard v1 - GSM8K | 0 | Accuracy (%) (5-shot) | 4.5 |
Interactive version: theaggregate.ai/model?slug=nebula-xb-v1-0 · How It Works · Data refreshed daily, snapshot 2026-11-08.