List your company

Math on HMMT25 (Avg@16)

44.5Avg@16TRD
27.23631.71836.240.682Jun 7, 2026

Evaluation results

MethodLinks
TRDBackbone=Qwen3-8B (w/ thinking), Traj.=yr2026.0644.5
+Reverse KL w/ Top-KBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0644.4
+Reverse KLBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0644.3
+Forward KL w/ ClipBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0642.5
BaseBackbone=Qwen3-8B (w/ thinking), Traj.=–2026.0641.5
+Forward KLBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0640.3
TRDBackbone=Qwen3-4B-Instruct-2507, Traj.=yr2026.0632.1
BaseBackbone=Qwen3-4B-Instruct-2507, Traj.=–2026.0631.3
+Reverse KL w/ Top-KBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.0631
+Forward KLBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.0630.8
+Reverse KLBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.0630
+Forward KL w/ ClipBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.0627.9