Jev مقابل التعلم الآلي التقليدي: اختبارات على 8 مجموعات بيانات

I put Jev against 11 classical ML pipelines across 8 datasets. The result wasn't "LLMs beat traditional ML." On IMDb, Jev Zero-Shot hit 96.3% vs 88.4% for the best classical pipeline. On Banking77, a Linear SVM won: 89.7% vs 78.9%. And across all 4 tabular datasets, classical