naps-llama-3.1-8B-instruct-v02 (NAPS-ai): benchmark results

Provider: Other.

Aggregate Rating 1428 ± 43, rank #4939 of 10552 rated models, from 9 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open Korean LLM - Ko-GSM8K (flexible extract)55.12Exact match, flexible extract (%)89.1
Open Korean LLM - Ko-IFEval39.95Strict accuracy, prompt/instruction mean (%)81.5
Open Korean LLM - Ko-EQ-Bench26.05EQ-Bench score72.6
Open Korean LLM Leaderboard39.35Average Score (%)71.2
Open Korean LLM - Ko-GPQA-Diamond24.24Normalized accuracy (%)62.4
Open Korean LLM - KorNAT-Harmless63.35Normalized accuracy (%)60.9
Open Korean LLM - KorNAT-Helpful46.43Normalized accuracy (%)46
Open Korean LLM - KorNAT-CKA22.85Normalized accuracy (%)41.4
Open Korean LLM - Ko-Winogrande56.91Accuracy (%)28.5

Interactive version: theaggregate.ai/model?slug=naps-llama-3-1-8b-instruct-v02-naps-ai · How It Works · Data refreshed daily, snapshot 2026-11-16.