List your company

Mathematics on AMC 24 (avg@32 accuracy)

46Accuracy (avg@32)IBTPO
13.86422.20730.5538.893May 27, 2026

Evaluation results

MethodLinks
IBTPOBase Model=Qwen3-8B-Base2026.0546
TreeRLBase Model=Qwen3-8B-Base2026.0540.5
TreePOBase Model=Qwen3-8B-Base2026.0540.1
IBROBase Model=Qwen3-8B-Base2026.0539.5
Vanilla GRPOBase Model=Qwen3-8B-Base2026.0539.4
GRPO w/ Entropy Reg†Base Model=Qwen3-8B-Base2026.0539.4
GRPO w/ Clip-higherBase Model=Qwen3-8B-Base2026.0538.9
Initial ModelBase Model=Qwen3-8B-Base2026.0526.8
IBTPOBase Model=Qwen3-1.7B-Base2026.0523.4
IBROBase Model=Qwen3-1.7B-Base2026.0520.8
TreeRLBase Model=Qwen3-1.7B-Base2026.0520.6
GRPO w/ Clip-higherBase Model=Qwen3-1.7B-Base2026.0520.5
Vanilla GRPOBase Model=Qwen3-1.7B-Base2026.0519.7
TreePOBase Model=Qwen3-1.7B-Base2026.0519.5
GRPO w/ Entropy Reg†Base Model=Qwen3-1.7B-Base2026.0518.8
Initial ModelBase Model=Qwen3-1.7B-Base2026.0515.1