List your company

Mathematical Reasoning on AMOBench (avg@8)

16Avg@8 ScoreSRT
0.7124.6818.6512.619Apr 13, 2026

Evaluation results

MethodLinks
SRTBase Model=Qwen3-4B-Instruct2026.0416
SD-ZEROBase Model=Qwen3-4B-Instruct2026.0416
RFTBase Model=Qwen3-4B-Instruct2026.0411.3
GRPOBase Model=Qwen3-4B-Instruct2026.0411
Qwen3-4B-InstructBase Model=Qwen3-4B-Instruct2026.049.8
SDFTBase Model=Qwen3-4B-Instruct2026.049
SFTBase Model=Qwen3-4B-Instruct2026.047.3
SD-ZEROBase Model=Olmo-3-7B-Instruct2026.045.5
GRPOBase Model=Olmo-3-7B-Instruct2026.044.8
SRTBase Model=Olmo-3-7B-Instruct2026.043.5
RFTBase Model=Olmo-3-7B-Instruct2026.042.3
Olmo-3-7B-InstructBase Model=Olmo-3-7B-Instruct2026.041.3
SFTBase Model=Olmo-3-7B-Instruct2026.041.3
SDFTBase Model=Olmo-3-7B-Instruct2026.041.3