Commonsense Reasoning on Hebrew Winograd Schema Challenge
83.45AccuracyLlama-3.3-70B-Instruct
Evaluation Results
| Method | Links | |
|---|---|---|
| Llama-3.3-70B-InstructSize Category=Large2026.02 | 83.45 | |
| aya-expanse-32BSize Category=Large2026.02 | 80.58 | |
| gemma3-27B-itSize Category=Large2026.02 | 79.86 | |
| DictaLM-3.0-24B-ThinkingSize Category=Large, Reasoning Mode=Thinking2026.02 | 78.06 | |
| gemma-3-12b-itSize Category=Smaller (~12B)2026.02 | 75.9 | |
| DictaLM-3.0-Nemotron-12B-InstructSize Category=Smaller (~12B)2026.02 | 73.74 | |
| Qwen3-14B (think)Size Category=Smaller (~12B), Reasoning Mode=Thinking2026.02 | 73.38 | |
| DictaLM-3.0-1.7B-InstructSize Category=Tiny (~1.7B)2026.02 | 58.2 | |
| DictaLM-3.0-1.7B-ThinkingSize Category=Tiny (~1.7B), Reasoning Mode=Thinking2026.02 | 55.76 | |
| Qwen3-1.7B (think)Size Category=Tiny (~1.7B), Reasoning Mode=Thinking2026.02 | 51.08 | |
| gemma-3-1b-itSize Category=Tiny (~1.7B)2026.02 | 47.84 |