Gembo-v1.1-70B: benchmark results
ChuckMcSneed's Gembo merge built from Sao10K/Euryale-1.3-L2-70B with multiple LoRA-derived additions described in its card. The published full-weight checkpoint is distinct from quantized conversions and from its input releases. Version 1.1 changes the mixture relative to v1; its card describes adding more human-data-derived material without establishing a measured training-token count. Provider: ChuckMcSneed. Access: Open.
Aggregate Rating 1466 ± 34, rank #3015 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 70.63 | Accuracy (%) (5-shot) | 94.2 |
| Open LLM Leaderboard v1 - ARC Challenge | 70.99 | Normalized accuracy (%) (25-shot) | 87.5 |
| Open LLM Leaderboard v1 - HellaSwag | 86.9 | Normalized accuracy (%) (10-shot) | 83.6 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 62.45 | MC2 (%) (0-shot) | 79.1 |
| Open LLM Leaderboard v1 - WinoGrande | 80.51 | Accuracy (%) (5-shot) | 74.3 |
| Open LLM Leaderboard v1 - GSM8K | 50.64 | Accuracy (%) (5-shot) | 63 |
Interactive version: theaggregate.ai/model?slug=gembo-v1-1-70b · How It Works · Data refreshed daily, snapshot 2026-11-08.