L-MChat-Small: benchmark results
Artples' L-MChat Small, a SLERP merge of phi-2-orange-v2 and Einstein-v4-phi2, testing how small sub-3B merges perform. Provider: Artples. Released 2024-04-11. Access: Open.
Aggregate Rating 1307 ± 26, rank #1466 of 1656 rated models, from 20 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 58.98 | Accuracy (%) (5-shot) | 70.7 |
| Open LLM Leaderboard v1 - ARC Challenge | 61.6 | Normalized accuracy (%) (25-shot) | 53.5 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 49.94 | MC2 (%) (0-shot) | 46.6 |
| Open LLM Leaderboard - BBH | 48.23 | Score | 42.1 |
| Open LLM Leaderboard v1 - MMLU | 57.41 | Accuracy (%) (5-shot) | 41.1 |
| Open LLM Leaderboard v1 - WinoGrande | 74.98 | Accuracy (%) (5-shot) | 39.7 |
| Open LLM Leaderboard - IFEval | 32.87 | Score | 30.8 |
| Open LLM Leaderboard v1 - HellaSwag | 75.9 | Normalized accuracy (%) (10-shot) | 27.6 |
| Open LLM Leaderboard - MMLU-Pro | 24.64 | Score | 25.2 |
| Open LLM Leaderboard - MuSR | 36.96 | Score | 23.6 |
| Open LLM Leaderboard - GPQA | 26.76 | Score | 22 |
| Open LLM Leaderboard - MATH Level 5 | 3.78 | Score | 21.8 |
Interactive version: theaggregate.ai/model?slug=l-mchat-small · How It Works · Data refreshed daily, snapshot 2026-10-20.