ResearchBenchmarksCommonsense Reasoning on Hellaswag EU20 (test)Follow48.83AccuracyEuroLM35.143638.696842.2545.8032Dec 25, 2025Evaluation ResultsMethodMethodLinksAccuracyEuroLM2025.1248.83Gamayun2025.1246.67Gemma32025.1242.1Qwen32025.1239.5Qwen2.52025.1239.01Llama3.22025.1235.67