List your company
ResearchFormerly WizWand

Mathematical Reasoning on AIME'25 (accuracy %)

75.83AIME'25 AccuracyMOPD
-2.617217.748938.11558.4811Feb 10, 2026Mar 4, 2026Mar 27, 2026Apr 18, 2026May 11, 2026Jun 2, 2026Jun 25, 2026
Updated 4mo ago

Evaluation results

MethodLinks
MOPDinference_mode=thinking mode2026.0675.83
Teacher Bestinference_mode=thinking mode2026.0675.42
NebulaExp-reasoning-SFTBackbone=Qwen3-8B-base2026.0672.08
Base (SFT)inference_mode=thinking mode2026.0670.83
Qwen3-14BBackbone=Qwen3-14B2026.0670.4
QWQ-32BBackbone=QWQ-32B2026.0669.5
Qwen3-8B-thinkingBackbone=Qwen3-8B, Mode=thinking2026.0665.71
Qwen3-8Binference_mode=thinking mode2026.0665.71
LaPha / sg@128Base model=Qwen2.5-7B, Tool=✓2026.0253.3
LaPha / sg@128Base model=Qwen2.5-Math-7B, Tool=✓2026.0253.3
GPT-o1-miniTool=✗2026.0251.7
LaPha / sg@128Base model=Qwen2.5-Math-1.5B, Tool=✓2026.0243.3
ToRLBase model=Qwen2.5-Math-7B, Tool=✓2026.0230
ToRLBase model=Qwen2.5-Math-1.5B, Tool=✓2026.0226.7
LaPhaBase model=Qwen2.5-Math-1.5B, Tool=✓2026.0226.7
DAPOBase model=Qwen2.5-Math-7B, Tool=✗2026.0223.3
LaPha / sg@128Base model=Qwen2.5-1.5B, Tool=✓2026.0216.7
DAPOBase model=Qwen2.5-7B, Tool=✗2026.0216.7
SFTBase model=Qwen2.5-Math-7B, Tool=✗2026.0216.7
SFTBase model=Qwen2.5-Math-7B, Tool=✓2026.0216.7
DAPOBase model=Qwen2.5-Math-1.5B, Tool=✗2026.0214.4
SFTBase model=Qwen2.5-Math-1.5B, Tool=✓2026.0213.3
PrimeBase model=Qwen2.5-Math-7B, Tool=✗2026.0213.3
LaPhaBase model=Qwen2.5-1.5B, Tool=✓2026.0210.4
SFTBase model=Qwen2.5-Math-1.5B, Tool=✗2026.0210
DAPOBase model=Qwen2.5-1.5B, Tool=✗2026.026.7
SFTBase model=Qwen2.5-7B, Tool=✗2026.026.7
SimpleRLBase model=Qwen2.5-Math-7B, Tool=✗2026.026.7
GPT-4oTool=✗2026.025.8
SFTBase model=Qwen2.5-1.5B, Tool=✗2026.020.4