List your company

Math on AMOBench (Avg@16)

17.3Avg@16TRD
8.9811.1413.315.46Jun 7, 2026

Evaluation results

MethodLinks
TRDBackbone=Qwen3-8B (w/ thinking), Traj.=yr2026.0617.3
+Reverse KLBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0616.8
BaseBackbone=Qwen3-8B (w/ thinking), Traj.=–2026.0615.9
+Forward KL w/ ClipBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0615.9
+Forward KLBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0615.7
+Reverse KL w/ Top-KBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0615.2
BaseBackbone=Qwen3-4B-Instruct-2507, Traj.=–2026.0610.6
TRDBackbone=Qwen3-4B-Instruct-2507, Traj.=yr2026.0610.3
+Reverse KLBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.0610.1
+Reverse KL w/ Top-KBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.069.6
+Forward KLBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.069.5
+Forward KL w/ ClipBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.069.3