List your company

Mathematical Reasoning on Average (AIME24, AIME25, MATH500, AMC23, Hmmt25, Olympiad)

30.07Average PerformanceCARE
15.031618.935822.8426.7442May 25, 2026

Evaluation results

MethodLinks
CAREBackbone=Phi4-mini-instruct2026.0530.07
GTBackbone=Phi4-mini-instruct2026.0530.06
OracleBackbone=Phi4-mini-instruct2026.0527.6
TTRLBackbone=Phi4-mini-instruct2026.0526.32
ProbBackbone=Phi4-mini-instruct2026.0526.25
RandomBackbone=Phi4-mini-instruct2026.0526.23
EntropyBackbone=Phi4-mini-instruct2026.0525.48
VanillaBackbone=Phi4-mini-instruct2026.0515.61