List your company

Mathematical Reasoning on DEEPMATH-103K

95.5AccuracyTRS
58.16467.85777.5587.243Apr 23, 2026

Evaluation results

MethodLinks
TRSModel=Gemini-3-Flash2026.0495.56,10682.5
DirectModel=Doubao Seed (Seed-1.8 Pro)2026.0495.13,453100
TRSModel=Doubao Seed (Seed-1.8 Pro)2026.0494.91,41146.2
DirectModel=Gemini-3-Flash2026.0494.87,490100
DirectModel=GPT-OSS-120B2026.0493.71,257100
TRSModel=GPT-OSS-120B2026.0493.797683.1
TRSModel=GPT-4o-mini2026.0461.465093.1
DirectModel=GPT-4o-mini2026.0459.6819100