List your company
ResearchFormerly WizWand

Mathematical Reasoning on AIME 2024-2025 (Accuracy (pass@1))

46.67Accuracy (pass@1)BF16
-1.866810.734123.33535.9359May 27, 2025Jul 20, 2025Sep 13, 2025Nov 7, 2025Dec 31, 2025Feb 24, 2026Apr 20, 2026
Updated 6mo ago

Evaluation results

MethodLinks
BF16Model=DeepSeek-R1-Distill-Qwen-7B, Bit-width=BF16, Maximum number of generation tokens=32,7682025.1246.67
R1-CI-14B2025.0542
MixKVQModel=DeepSeek-R1-Distill-Qwen-7B, Bit-width=C3.4, Maximum number of generation tokens=32,7682025.1240
KIVIModel=DeepSeek-R1-Distill-Qwen-7B, Bit-width=KV4, Maximum number of generation tokens=32,7682025.1236.67
KVQuantModel=DeepSeek-R1-Distill-Qwen-7B, Bit-width=KV4, Maximum number of generation tokens=32,7682025.1233.33
RotateKVModel=DeepSeek-R1-Distill-Qwen-7B, Bit-width=KV4, Maximum number of generation tokens=32,7682025.1233.33
Qwen-2.5-14BInference Mode=CodeSteer2025.0533.3
KVTunerModel=DeepSeek-R1-Distill-Qwen-7B, Bit-width=C3.92, Maximum number of generation tokens=32,7682025.1231.67
KIVIModel=DeepSeek-R1-Distill-Qwen-7B, Bit-width=KV2, Maximum number of generation tokens=32,7682025.1230
Qwen-2.5-14BInference Mode=All Text2025.0530
R1-CI-7B2025.0515
Qwen-2.5-7BInference Mode=All Text2025.058.33
Qwen-2.5-7BInference Mode=CodeSteer2025.058.33
Standard AR2026.048.3
Best-of-16sampling=Best-of-162026.048.3
LPSR2026.048.3
CoCoNuT2026.046.7
STIR-Static2026.041.7
KVQuantModel=DeepSeek-R1-Distill-Qwen-7B, Bit-width=KV2, Maximum number of generation tokens=32,7682025.121.67
Prompted SC2026.040