Skip to content

Benchmark

Bridging the Gap Winogrande Amharic (am) human-translation minus machine-translation difference

Multilingual benchmarks compare model performance across languages.

Models measured
6
Scores recorded
6
Sources
1
IQ weight
No weight published

Score distribution

0.617
Mean
0.4
Median
1
Std dev
-0.1
10th percentile
2.5
90th percentile
-0.5
Min
-0.52.5
The band covers the middle 80% of scores. The line marks the median.

Results

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

ModelScore
GPT 4otable: A.11, section: machine-translation-difference, task: Winogrande, language: am, shots: 5· Bridging the Gap (arXiv:2412.12417v1) data2.5
GPT-3.5 Turbotable: A.11, section: machine-translation-difference, task: Winogrande, language: am, shots: 5· Bridging the Gap (arXiv:2412.12417v1) data1.3
GPT-4 Turbotable: A.11, section: machine-translation-difference, task: Winogrande, language: am, shots: 5· Bridging the Gap (arXiv:2412.12417v1) data0.1
Llama 3 70B Instructtable: A.11, section: machine-translation-difference, task: Winogrande, language: am, shots: 5· Bridging the Gap (arXiv:2412.12417v1) data0.4
Llama 3 8B Instructtable: A.11, section: machine-translation-difference, task: Winogrande, language: am, shots: 5· Bridging the Gap (arXiv:2412.12417v1) data-0.1
Phi-3 Mini 4K Instructtable: A.11, section: machine-translation-difference, task: Winogrande, language: am, shots: 5· Bridging the Gap (arXiv:2412.12417v1) data-0.5

Data credits: Tuka Alhanai, Adam Kasumovic, Mohammad Ghassemi, Aven Zitzelberger, Jessica Lundin and Guillaume Chabot-Couture. Bridging the Gap: Enhancing LLM Performance for Low-Resource African Languages with New Benchmarks, Fine-Tuning, and Cultural Adjustments, arXiv:2412.12417v1. CC BY 4.0 article numeric/protocol extract, restructured with values unchanged. No task or model rights asserted. (licence)