llm-jp-3-980m-instruct2: benchmark results
LLM-jp 3 instruction-tuned release developed by NII's LLM-jp project, using Llama architecture and a 4,096-token context window. The named instruction-tuning generation remains distinct from the pretrained checkpoint and other instruct generations. Provider: National Institute of Informatics. Access: Open.
Aggregate Rating 1287 ± 52, rank #8809 of 10556 rated models, from 62 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Japanese LLM - Xlsum JA Rouge1 | 27.39 | Score (%) | 63.9 |
| Open Japanese LLM - Xlsum JA RougeLsum | 22.97 | Score (%) | 61.2 |
| pfgen-bench - Completion Mode - Truthfulness | 0.72 | Truthfulness Score | 49 |
| pfgen-bench - Completion Mode - Fluency | 0.56 | Fluency Score | 48.2 |
| pfgen-bench - Completion Mode - Score | 0.46 | pfgen Score (mean of three) | 47 |
| Open Japanese LLM - Xlsum JA Bert Score JA F1 | 68.61 | Score (%) | 43.9 |
| pfgen-bench - Completion Mode - Helpfulness | 0.09 | Helpfulness Score | 43.1 |
| Open Japanese LLM - Xlsum JA Rouge2 | 9 | Score (%) | 40.9 |
| Open Japanese LLM - SUM | 9 | Score (%) | 40.7 |
| pfgen-bench - QA Mode - Fluency | 0.53 | Fluency Score | 38.2 |
| pfgen-bench - QA Mode - Score | 0.44 | pfgen Score (mean of three) | 36.4 |
| pfgen-bench - QA Mode - Helpfulness | 0.1 | Helpfulness Score | 35.2 |
Interactive version: theaggregate.ai/model?slug=llm-jp-3-980m-instruct2 · How It Works · Data refreshed daily, snapshot 2026-11-17.