SauerkrautLM-Gemma-2B: benchmark results
An early Gemma-2B fine-tune jointly released by VAGOsolutions and Hyperspace.ai using SFT and laser-QLoRA partial freezing, with credit to the LaserRMT research collaboration. This is the original Gemma generation, distinct from Gemma 2 Instruct releases. Provider: VAGOsolutions. Access: Open.
Aggregate Rating 1261 ± 32, rank #9158 of 10916 rated models, from 12 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 26.76 | Accuracy (%) (5-shot) | 44.8 |
| Open LLM Leaderboard v1 - ARC Challenge | 48.72 | Normalized accuracy (%) (25-shot) | 23.5 |
| Open LLM Leaderboard - MuSR | 36.76 | Score | 22.6 |
| Open LLM Leaderboard v1 - MMLU | 42.9 | Accuracy (%) (5-shot) | 22.2 |
| Open LLM Leaderboard v1 - WinoGrande | 67.96 | Accuracy (%) (5-shot) | 22 |
| Open LLM Leaderboard v1 - HellaSwag | 71.41 | Normalized accuracy (%) (10-shot) | 21.5 |
| Open LLM Leaderboard - IFEval | 24.75 | Score | 19.7 |
| Open LLM Leaderboard - BBH | 34.16 | Score | 17.8 |
| Open LLM Leaderboard - MATH Level 5 | 2.79 | Score | 17.8 |
| Open LLM Leaderboard - MMLU-Pro | 14.69 | Score | 12.2 |
| Open LLM Leaderboard - GPQA | 25.67 | Score | 9 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 35.77 | MC2 (%) (0-shot) | 2.7 |
Interactive version: theaggregate.ai/model?slug=sauerkrautlm-gemma-2b · How It Works · Data refreshed daily, snapshot 2026-11-08.