Linkbricks-Horizon-AI-Korean-llama3.1-sft-rlhf-dpo-8B: benchmark results

Provider: Other. Access: Open.

Unified ELO 1550 ± 17, rank #842 of 2656 rated models, from 10 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open Ko-LLM Leaderboard47.12Average (%)94.6
Open Korean LLM Leaderboard47.12Average Score (%)94.6
Open Korean LLM - Ko-IFEval51.13Strict accuracy, prompt/instruction mean (%)94.4
Open Korean LLM - Ko-GSM8K (flexible extract)59.36Exact match, flexible extract (%)93.3
Open Korean LLM - KorNAT-CKA32.54Normalized accuracy (%)89
Open Korean LLM - Ko-EQ-Bench35.39EQ-Bench score85.4
Open Korean LLM - Ko-GPQA-Diamond24.24Normalized accuracy (%)62.4
Open Korean LLM - Ko-Winogrande60.77Accuracy (%)57.3
Open Korean LLM - KorNAT-Helpful45.9Normalized accuracy (%)40.4
Open Korean LLM - KorNAT-Harmless62.21Normalized accuracy (%)40.3

Interactive version: theaggregate.ai/model?slug=linkbricks-horizon-ai-korean-llama3-1-sft-rlhf-dpo-8b · How It Works · Data refreshed daily, snapshot 2026-09-19.