List your company

Logical Reasoning on MNLogic 5dgt

72.78AccuracyVLC
44.263251.666659.0766.4734Mar 25, 2026

Evaluation results

MethodLinks
VLCBackbone Scale=7B, Random Seeds=52026.0372.78
PrismBackbone Scale=7B, Random Seeds=52026.0350.19
End2end RSBackbone Scale=7B, Random Seeds=52026.0349.51
End2end FTBackbone Scale=7B, Random Seeds=52026.0346.55
ViperGPTBackbone Scale=7B, Random Seeds=52026.0345.36