List your company

Mathematical Reasoning on MATH-P

48.4Pass@1FPA
19.48826.99434.542.006Sep 24, 2025

Evaluation results

MethodLinks
FPABackbone=Qwen3-4B-Base2025.0948.4
FPABackbone=Qwen2.5-Math-7B2025.0944.6
Off-RLBackbone=Qwen2.5-Math-7B2025.0943.9
DPOBackbone=Qwen3-4B-Base2025.0943.7
A*-POBackbone=Qwen3-4B-Base2025.0942.6
A*-POBackbone=Qwen2.5-Math-7B2025.0942.3
Off-RLBackbone=Qwen3-4B-Base2025.0941.4
KTOBackbone=Qwen3-4B-Base2025.0941.1
SFTBackbone=Qwen2.5-Math-7B2025.0940.1
BaseBackbone=Qwen2.5-Math-7B2025.0939.8
RPOBackbone=Qwen3-4B-Base2025.0939
RPOBackbone=Qwen2.5-Math-7B2025.0938.4
KTOBackbone=Qwen2.5-Math-7B2025.0938.4
DPOBackbone=Qwen2.5-Math-7B2025.0938.1
SFTBackbone=Qwen3-4B-Base2025.0934.6
BaseBackbone=Qwen3-4B-Base2025.0933.2
A*-POBackbone=Llama-3.2-3B-Instruct2025.0921.6
RPOBackbone=Llama-3.2-3B-Instruct2025.0921.4
DPOBackbone=Llama-3.2-3B-Instruct2025.0921.3
KTOBackbone=Llama-3.2-3B-Instruct2025.0921.2
Off-RLBackbone=Llama-3.2-3B-Instruct2025.0921.1
BaseBackbone=Llama-3.2-3B-Instruct2025.0920.9
FPABackbone=Llama-3.2-3B-Instruct2025.0920.7
SFTBackbone=Llama-3.2-3B-Instruct2025.0920.6