List your company

Commonsense Reasoning on HellaSwag (Relative Improvement Analysis)

3.81Average Relative ImprovementTBDF
-0.11080.90711.9252.9429Jan 29, 2026

Evaluation results

MethodLinks
TBDFFiltering Mode=General, Models=Gemma 3 (270M, 1B, 4B)2026.013.81—
TBDFFiltering Mode=FW-EDU, Models=Gemma 3 (270M, 1B, 4B)2026.012.36—
CBFiltering Mode=FW-EDU, Models=Gemma 3 (270M, 1B, 4B)2026.010.04—