List your company

Math on AIME24 (Avg@16)

76.5Avg@16Base
28.015240.602653.1965.7774May 8, 2026May 13, 2026May 18, 2026May 23, 2026May 28, 2026Jun 2, 2026Jun 7, 2026

Evaluation results

MethodLinks
BaseBackbone=Qwen3-8B (w/ thinking), Traj.=–2026.0676.5
+Forward KL w/ ClipBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0676.5
TRDBackbone=Qwen3-8B (w/ thinking), Traj.=yr2026.0676.5
+Reverse KL w/ Top-KBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0675.6
+Reverse KLBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0675.4
+Forward KLBackbone=Qwen3-8B (w/ thinking), Traj.=yo2026.0674.8
BaseBackbone=Qwen3-4B-Instruct-2507, Traj.=–2026.0663.1
TRDBackbone=Qwen3-4B-Instruct-2507, Traj.=yr2026.0663.1
+Reverse KL w/ Top-KBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.0663
+Forward KL w/ ClipBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.0662.1
+Forward KLBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.0660.3
+Reverse KLBackbone=Qwen3-4B-Instruct-2507, Traj.=yo2026.0658.4
AIPOModel=Qwen2.5-32B-Instruct2026.0539.06
GRPOModel=Qwen2.5-32B-Instruct2026.0535.71
SFTModel=Qwen2.5-32B-Instruct2026.0532.61
OriginalModel=Qwen2.5-32B-Instruct2026.0529.88