List your company

Mathematical Reasoning on OlympiadBench (Accuracy, Thinking Token Usage (%))

71.5AccuracyAT
48.93254.79160.6566.509Jun 22, 2026

Evaluation results

MethodLinks
ATModel Scale=8B, Thinking Strategy=AT2026.0671.5—
DARTModel Scale=8B, Thinking Strategy=DART2026.0669.845
DARTModel Scale=14B, Thinking Strategy=DART2026.066215
DARTModel Scale=32B, Thinking Strategy=DART2026.0658.516
ATModel Scale=32B, Thinking Strategy=AT2026.0654—
ATModel Scale=14B, Thinking Strategy=AT2026.0653—
NTModel Scale=14B, Thinking Strategy=NT2026.0651.5—
NTModel Scale=32B, Thinking Strategy=NT2026.0650.5—
NTModel Scale=8B, Thinking Strategy=NT2026.0649.8—