Linkbricks-Horizon-AI-Llama-3.3-Japanese-70B-sft-dpo-base: benchmark results
Provider: Linkbricks Horizon-AI. Access: Open.
Aggregate Rating 1525 ± 35, rank #2015 of 10916 rated models, from 50 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Japanese LLM - MMLU EN Exact Match | 80.36 | Score (%) | 93.3 |
| Open Japanese LLM - ALT E TO J Bert Score JA F1 | 87.06 | Score (%) | 90.9 |
| Open Japanese LLM - Janli Exact Match | 86.94 | Score (%) | 90 |
| Open Japanese LLM - Wikicorpus E TO J Bleu JA | 15.18 | Score (%) | 89.2 |
| Open Japanese LLM - Wikicorpus J TO E Bleu EN | 12.17 | Score (%) | 87.7 |
| Open Japanese LLM - ALT E TO J Bleu JA | 13.66 | Score (%) | 86.5 |
| Open Japanese LLM - RC | 91.78 | Score (%) | 86.1 |
| Open Japanese LLM - ALT J TO E Bert Score EN F1 | 95.64 | Score (%) | 85.8 |
| Open Japanese LLM - MR | 93.6 | Score (%) | 85.8 |
| Open Japanese LLM - Jsem Exact Match | 79.92 | Score (%) | 85 |
| Open Japanese LLM - Niilc Char F1 | 57.21 | Score (%) | 84.4 |
| Open Japanese LLM - QA | 61.51 | Score (%) | 84.1 |
Interactive version: theaggregate.ai/model?slug=linkbricks-horizon-ai-llama-3-3-japanese-70b-sft-dpo-base · How It Works · Data refreshed daily, snapshot 2026-11-08.