List your company

Mathematical Reasoning on AMC (Mean Accuracy)

59.5Mean AccuracyDiSCTT
10.30823.07935.8548.621Mar 5, 2026

Evaluation results

MethodLinks
DiSCTTModel=Qwen-2.5-7B-Instruct2026.0359.5
DiSCTTModel=Qwen-3-4B-Base2026.0357
EVOL-RLModel=Qwen-2.5-7B-Instruct2026.0355
EVOL-RLModel=Qwen-3-4B-Base2026.0351.8
TTRLModel=Qwen-2.5-7B-Instruct2026.0351.1
DiSCTTModel=Qwen-3-1.7B-Base2026.0348.8
TTRLModel=Qwen-3-4B-Base2026.0346.6
EVOL-RLModel=Qwen-3-1.7B-Base2026.0344.9
TTRLModel=Qwen-3-1.7B-Base2026.0340.2
BaseModel=Qwen-2.5-7B-Instruct2026.0339.6
DiSCTTModel=LLaMA-3.2-3B-Instruct2026.0339.1
BaseModel=Qwen-3-4B-Base2026.0338.6
EVOL-RLModel=LLaMA-3.2-3B-Instruct2026.0332.2
DiSCTTModel=Qwen-2.5-0.5B-Instruct2026.0327.7
TTRLModel=LLaMA-3.2-3B-Instruct2026.0327.2
DiSCTTModel=LLaMA-3.2-1B-Instruct2026.0326.4
EVOL-RLModel=LLaMA-3.2-1B-Instruct2026.0325.4
EVOL-RLModel=Qwen-2.5-0.5B-Instruct2026.0323.1
TTRLModel=LLaMA-3.2-1B-Instruct2026.0322.5
TTRLModel=Qwen-2.5-0.5B-Instruct2026.0320.3
BaseModel=Qwen-3-1.7B-Base2026.0319.8
BaseModel=LLaMA-3.2-1B-Instruct2026.0315.4
BaseModel=LLaMA-3.2-3B-Instruct2026.0312.5
BaseModel=Qwen-2.5-0.5B-Instruct2026.0312.2