List your company

Mathematical Reasoning on AIME In-Distribution 2026 (Accuracy)

70AccuracySkillFlow
31.863241.764151.66561.5659May 13, 2026

Evaluation results

MethodLinks
SkillFlowMethod Category=Ours2026.0570
FlowSteerMethod Category=Agent+RL2026.0563.33
AgentFlowMethod Category=Agent+RL2026.0560
SkillRLMethod Category=Agent+RL2026.0556.67
Qwen3.5-AFlowMethod Category=AFlow, Backbone=Qwen3.52026.0553.33
v4-flashMethod Category=Baseline2026.0550
Qwen3.5-9BMethod Category=Baseline2026.0546.67
Qwen3.5-SFTMethod Category=SFT, Backbone=Qwen3.52026.0536.67
Qwen3.5-GRPOMethod Category=GRPO, Backbone=Qwen3.52026.0533.33