List your company

Scientific Reasoning on PHYBench

30.34AccuracyInftyThink+
15.686419.490723.29527.0993Feb 6, 2026

Evaluation results

MethodLinks
InftyThink+RL Setting=Task reward only (✓ T), Sampling Temperature=0.7, Number of Samples=322026.0230.3430.34219.01
InftyThink+RL Setting=Task and efficiency reward (✓ T+E), Sampling Temperature=0.7, Number of Samples=322026.0229.6214.6397.57
VanillaRL Setting=Task reward only (✓ T), Sampling Temperature=0.7, Number of Samples=322026.0220.1120.46320.06
InftyThink+RL Setting=Cold start only (✗), Sampling Temperature=0.7, Number of Samples=322026.0219.1414.694.55
VanillaRL Setting=Cold start only (✗), Sampling Temperature=0.7, Number of Samples=322026.0216.2513.44149.38