Llama 3 10B IT Kor Extented Chang Pro8: benchmark results
Provider: Meta. Access: Open.
Aggregate Rating 1409 ± 27, rank #5653 of 10916 rated models, from 15 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GPQA-Diamond | 25.76 | Normalized accuracy (%) | 81.3 |
| Open Korean LLM - Ko-EQ-Bench | 30.53 | EQ-Bench score | 79.3 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 40.79 | Exact match, flexible extract (%) | 75.2 |
| Open Korean LLM Leaderboard | 39.77 | Average Score (%) | 74 |
| Open LLM Leaderboard v1 - GSM8K | 56.79 | Accuracy (%) (5-shot) | 68.6 |
| Open LLM Leaderboard v1 - MMLU | 64.31 | Accuracy (%) (5-shot) | 68 |
| Open Korean LLM - Ko-Winogrande | 61.17 | Accuracy (%) | 60.3 |
| Open LLM Leaderboard v1 - WinoGrande | 77.27 | Accuracy (%) (5-shot) | 54.6 |
| Open Korean LLM - KorNAT-CKA | 24.73 | Normalized accuracy (%) | 53 |
| Open LLM Leaderboard v1 - ARC Challenge | 57.85 | Normalized accuracy (%) (25-shot) | 38.3 |
| Open LLM Leaderboard v1 - HellaSwag | 79.34 | Normalized accuracy (%) (10-shot) | 37.5 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 46.99 | MC2 (%) (0-shot) | 36.1 |
Interactive version: theaggregate.ai/model?slug=llama-3-10b-it-kor-extented-chang-pro8 · How It Works · Data refreshed daily, snapshot 2026-11-08.