MixtureofMerges-MoE-2x7B-SLERPv0.9: benchmark results
jsfs11 release formed by SLERP merging jsfs11/MixtureofMerges-MoE-2x7b-v7 and jsfs11/MixtureofMerges-MoE-2x7bRP-v8. The recipe uses distinct attention and MLP interpolation schedules. Its Mixtral configuration retains 2 experts and selects 2 per token; it is distinct from both input releases. Provider: jsfs11. Access: Open.
Aggregate Rating 1465 ± 41, rank #3044 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - ARC Challenge | 73.12 | Normalized accuracy (%) (25-shot) | 95.8 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 74.83 | MC2 (%) (0-shot) | 95.2 |
| Open LLM Leaderboard v1 - HellaSwag | 88.76 | Normalized accuracy (%) (10-shot) | 93.9 |
| Open LLM Leaderboard v1 - WinoGrande | 83.58 | Accuracy (%) (5-shot) | 89.7 |
| Open LLM Leaderboard v1 - GSM8K | 69.22 | Accuracy (%) (5-shot) | 88.4 |
| Open LLM Leaderboard v1 - MMLU | 65 | Accuracy (%) (5-shot) | 80 |
Interactive version: theaggregate.ai/model?slug=mixtureofmerges-moe-2x7b-slerpv0-9 · How It Works · Data refreshed daily, snapshot 2026-11-08.