Llama-3-Instruct-8B-KTO-v0.2: benchmark results
Princeton NLP's Llama-3-Instruct-8B-KTO-v0.2, a Llama 3 8B Instruct model aligned via KTO, released alongside Princeton's SimPO preference-optimization paper. Provider: Princeton NLP. Released 2024-07-06. Access: Open.
Aggregate Rating 1408 ± 22, rank #1194 of 1654 rated models, from 20 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GSM8K (flexible extract) | 54.97 | Exact match, flexible extract (%) | 88.8 |
| Open LLM Leaderboard - IFEval | 72.9 | Score | 86.9 |
| Open Korean LLM - KorNAT-Helpful | 50.2 | Normalized accuracy (%) | 84.6 |
| Open Korean LLM - Ko-IFEval | 41.32 | Strict accuracy, prompt/instruction mean (%) | 84.4 |
| Open Korean LLM - Ko-GPQA-Diamond | 25.76 | Normalized accuracy (%) | 81.3 |
| Open Portuguese LLM - OAB Exams | 49.93 | Accuracy (%) | 75.2 |
| Open Portuguese LLM - ENEM | 68.93 | Accuracy (%) | 71.5 |
| Open Portuguese LLM - BLUEX | 57.3 | Accuracy (%) | 69.4 |
| Open Korean LLM Leaderboard | 37.45 | Average Score (%) | 61.9 |
| Open Portuguese LLM - ASSIN2 RTE | 90.91 | Macro F1 (%) | 61.4 |
| Open LLM Leaderboard - MMLU-Pro | 36.68 | Score | 58 |
| Open Korean LLM - KorNAT-CKA | 25.13 | Normalized accuracy (%) | 55.2 |
Interactive version: theaggregate.ai/model?slug=llama-3-instruct-8b-kto-v0-2 · How It Works · Data refreshed daily, snapshot 2026-10-24.