List your company

Mathematical Reasoning on HMMT Nov 25

73.9AVG@KCE + Distillation
24.537.32550.1562.975May 25, 2026

Evaluation results

MethodLinks
CE + DistillationBase Model=Nanbeige4-3B-2511-Thinking2026.0573.9
CEBase Model=Nanbeige4-3B-2511-Thinking2026.0572.8
SAPO + DistillationBase Model=Nanbeige4-3B-2511-Thinking2026.0572
SAPOBase Model=Nanbeige4-3B-2511-Thinking2026.0571.1
PPOBase Model=Nanbeige4-3B-2511-Thinking2026.0570.5
CEBase Model=Qwen3-4B-Thinking-25072026.0570.4
CE + DistillationBase Model=Qwen3-4B-Thinking-25072026.0569.9
GRPOBase Model=Nanbeige4-3B-2511-Thinking2026.0568.9
MSE + DistillationBase Model=Nanbeige4-3B-2511-Thinking2026.0568.1
SAPO + DistillationBase Model=Qwen3-4B-Thinking-25072026.0568
MSE + DistillationBase Model=Qwen3-4B-Thinking-25072026.0568
BaseBase Model=Nanbeige4-3B-2511-Thinking2026.0567.3
SAPOBase Model=Qwen3-4B-Thinking-25072026.0566.8
BaseBase Model=Qwen3-4B-Thinking-25072026.0566.6
GRPO + DistillationBase Model=Nanbeige4-3B-2511-Thinking2026.0566.5
PPOBase Model=Qwen3-4B-Thinking-25072026.0566
GRPOBase Model=Qwen3-4B-Thinking-25072026.0565.7
GRPO + DistillationBase Model=Qwen3-4B-Thinking-25072026.0564.2
MSEBase Model=Qwen3-4B-Thinking-25072026.0540.8
MSEBase Model=Nanbeige4-3B-2511-Thinking2026.0526.4