List your company

Mathematical Reasoning on MATH 500 (Acc, TR)

94.8AccuracyDynasor
54.2464.7775.385.83May 17, 2026

Evaluation results

MethodLinks
DynasorModel=Qwen3-30B-A3B-Thinking2026.0594.8-4.3
DEERModel=Qwen3-30B-A3B-Thinking2026.0594.834.9
Full CoTModel=Qwen3-30B-A3B-Thinking2026.0594.40
PUMAModel=Qwen3-30B-A3B-Thinking2026.0594.228.8
Full CoTModel=Llama-3.1-Nemotron-Nano-8B2026.0593.60
PUMAModel=Llama-3.1-Nemotron-Nano-8B2026.0592.615.7
DEERModel=Llama-3.1-Nemotron-Nano-8B2026.0591.822.7
No-ThinkModel=Qwen3-30B-A3B-Thinking2026.0591.851.7
Plan&BudgetModel=Qwen3-30B-A3B-Thinking2026.0591.851.4
DynasorModel=Llama-3.1-Nemotron-Nano-8B2026.0591.4-9
DEERModel=DeepSeek-R1-Distill-Qwen-7B2026.0590.639.6
Full CoTModel=DeepSeek-R1-Distill-Qwen-7B2026.05900
PUMAModel=DeepSeek-R1-Distill-Qwen-7B2026.0589.638.6
CCoTModel=Qwen3-30B-A3B-Thinking2026.0589.254.6
Plan&BudgetModel=Llama-3.1-Nemotron-Nano-8B2026.05898.3
CCoTModel=Llama-3.1-Nemotron-Nano-8B2026.0586.634.5
CCoTModel=DeepSeek-R1-Distill-Qwen-7B2026.058560.9
Plan&BudgetModel=DeepSeek-R1-Distill-Qwen-7B2026.0584.648.5
DynasorModel=DeepSeek-R1-Distill-Qwen-7B2026.0584.230.6
CoDModel=Qwen3-30B-A3B-Thinking2026.0582.829.9
CoDModel=DeepSeek-R1-Distill-Qwen-7B2026.0580.459.9
No-ThinkModel=DeepSeek-R1-Distill-Qwen-7B2026.057980
CoDModel=Llama-3.1-Nemotron-Nano-8B2026.057930.1
No-ThinkModel=Llama-3.1-Nemotron-Nano-8B2026.0562.8-125.5
Ans. Conv.Model=DeepSeek-R1-Distill-Qwen-7B2026.0561.876.5
Ans. Conv.Model=Qwen3-30B-A3B-Thinking2026.055787.8
Ans. Conv.Model=Llama-3.1-Nemotron-Nano-8B2026.0555.876.3