Mistral 7B (5-shot)
Reported on 4 benchmarks across 3 tasks · 2 papers
Note: results are matched by exact model name. Different papers may use the same name for different model variants.
Methodology4 results
- Average (%)· uses extra data· 2024-01-0862.5best: 87 (GPT-4 o1(300b))
- Average (%)· uses extra data· 2024-01-0862.5best: 87 (GPT-4 o1(300b))
- Average (%)· 2023-10-1060.1best: 87 (GPT-4 o1(300b))
- Average (%)· 2023-10-1060.1best: 87 (GPT-4 o1(300b))
Natural Language Processing2 results
- EM· 2023-10-1028.8best: 64 (Atlas (full, Wiki-dec-2018 index))
- EM· 2023-10-1069.9best: 87.5 (Claude 2 (few-shot, k=5))