List your company

Common Sense Reasoning on HELLASWAG 1800

53.8AccuracyQwen3-4B
35.28840.09444.949.706Jun 2, 2026

Evaluation results

MethodLinks
Qwen3-4B2026.0653.8
Talkie-1930variant=base2026.0652.2
GPT-2 XL2026.0641.1
TYPEWRITERLMvariant=base2026.0639
GPT-1900variant=base2026.0636