List your company
ResearchFormerly WizWand

Mathematical Reasoning on Olympiads

70.22Pass@1SGPO
40.330448.090255.8563.6098May 16, 2025
Updated 7mo ago

Evaluation results

MethodLinks
SGPOBackbone=QwQ-32B, Judge Model=DeepSeek-V3-03242025.0570.22
GRPOBackbone=QwQ-32B2025.0569.48
BaselineBackbone=QwQ-32B2025.0568.74
GRPOBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0556
SGPOBackbone=DeepSeek-R1-Distill-Qwen-7B, Judge Model=DeepSeek-V3-03242025.0555.11
SGPOBackbone=DeepSeek-R1-Distill-Qwen-7B, Judge Model=QwQ-32B, Stability Parameters=w/o {beta,gamma}2025.0555.11
SGPOBackbone=DeepSeek-R1-Distill-Qwen-7B, Judge Model=o4-mini-04162025.0554.67
SGPOBackbone=DeepSeek-R1-Distill-Qwen-7B, Judge Model=Qwen3-235B-A22B2025.0554.67
SGPOBackbone=DeepSeek-R1-Distill-Qwen-7B, Judge Model=QwQ-32B2025.0553.78
BaselineBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0549.93
SGPOBackbone=DeepSeek-R1-Distill-Llama-8B, Judge Model=Claude-3.72025.0548.44
SGPOBackbone=Qwen2.5-32B-Instruct, Judge Model=o4-mini-04162025.0546.81
GRPOBackbone=DeepSeek-R1-Distill-Llama-8B2025.0546.32
BaselineBackbone=Qwen2.5-32B-Instruct2025.0545.93
GRPOBackbone=Qwen2.5-32B-Instruct2025.0545.93
SGPOBackbone=Qwen2.5-14B-Instruct, Judge Model=o4-mini-04162025.0544
GRPOBackbone=Qwen2.5-14B-Instruct2025.0543.11
BaselineBackbone=Qwen2.5-14B-Instruct2025.0541.78
BaselineBackbone=DeepSeek-R1-Distill-Llama-8B2025.0541.48