List your company

Mathematical Reasoning on MATH 500 (Acc, FLOPs, A/C)

82.8AccuracyOurs
72.60875.25477.980.546May 7, 2026

Evaluation results

MethodLinks
OursExternal Models=false2026.0582.85.3415.5
RSDExternal Models=true2026.0579.84.5617.5
7B OnlyExternal Models=N/A2026.0579.69.18.75
STEERExternal Models=false2026.0579.66.3812.4
SpecReasonExternal Models=false2026.0577.47.7410
1.5B OnlyExternal Models=N/A2026.05731.8240.1