List your company

Mathematical Reasoning on AIME26 (accuracy)

44.06AIME26 AccuracyTOP-D (Ours)
4.633614.869325.10535.3407Jul 6, 2026

Evaluation results

MethodLinks
TOP-D (Ours)Student Model=Qwen3-8B-Base, Teacher Model=Qwen3-30B-A3B-Instruct-2507, Paradigm=Distillation, Evaluation Protocol=avg@322026.0744.06
DAPOStudent Model=Qwen3-8B-Base, Teacher Model=Qwen3-30B-A3B-Instruct-2507, Paradigm=RLVR, Evaluation Protocol=avg@322026.0732.29
OPDStudent Model=Qwen3-8B-Base, Teacher Model=Qwen3-30B-A3B-Instruct-2507, Paradigm=Distillation, Evaluation Protocol=avg@322026.0725.42
GRPOStudent Model=Qwen3-8B-Base, Teacher Model=Qwen3-30B-A3B-Instruct-2507, Paradigm=RLVR, Evaluation Protocol=avg@322026.0721.67
BaseStudent Model=Qwen3-8B-Base, Paradigm=-, Evaluation Protocol=avg@322026.076.15