koishi-instruct-3B: benchmark results
Provider: Other. Access: Open.
Aggregate Rating 1168 ± 38, rank #9777 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - ARC Challenge | 40.96 | Normalized accuracy (%) (25-shot) | 17.3 |
| Open LLM Leaderboard v1 - HellaSwag | 64.54 | Normalized accuracy (%) (10-shot) | 16.9 |
| Open LLM Leaderboard v1 - WinoGrande | 64.09 | Accuracy (%) (5-shot) | 16.9 |
| Open LLM Leaderboard v1 - GSM8K | 1.14 | Accuracy (%) (5-shot) | 15.7 |
| Open LLM Leaderboard v1 - MMLU | 26.58 | Accuracy (%) (5-shot) | 13 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 31.65 | MC2 (%) (0-shot) | 0.1 |
Interactive version: theaggregate.ai/model?slug=koishi-instruct-3b · How It Works · Data refreshed daily, snapshot 2026-11-08.