List your company
ResearchFormerly WizWand

Mathematical Reasoning on OlympiadBench Math (test)

35.9AccuracyTATA
2.41211.10619.828.494Jun 18, 2024Jul 28, 2024Sep 7, 2024Oct 18, 2024Nov 27, 2024Jan 7, 2025Feb 17, 2025
Updated 8mo ago

Evaluation results

MethodLinks
TATAModel=Qwen2.5Math-7B2025.0235.9
hybridModel=Qwen2.5-14B2025.0235.3
TATAModel=Qwen2.5-14B2025.0235.3
hybridModel=Qwen2.5Math-7B2025.0234.4
ensembleModel=Qwen2.5-14B2025.0232.3
hybridModel=Qwen2.5-7B2025.0231.7
TATAModel=Qwen2.5-7B2025.0231.1
ensembleModel=Qwen2.5Math-7B2025.0230.8
ensembleModel=Qwen2.5-7B2025.0230.2
ensembleModel=Qwen2.5Math-1.5B2025.0227.4
TATAModel=Qwen2.5Math-1.5B2025.0226.8
hybridModel=Qwen2.5Math-1.5B2025.0226.2
TATAModel=Qwen2.5-3B2025.0224.9
GPT-SelectModel=Qwen2.5Math-7B2025.0224.6
hybridModel=Qwen2.5-3B2025.0224.4
GPT-SelectModel=Qwen2.5-7B2025.0223.3
ensembleModel=Qwen2.5-3B2025.0223.1
GPT-SelectModel=Qwen2.5-14B2025.0223
DART-Math-DSMath-7B (Prop2Diff)number of samples=0.59M, Base Model=DeepSeekMath-7B, Sampling Strategy=Prop2Diff2024.0621.7
TATAModel=LLaMA-3-8B2025.0221.5
DART-Math-DSMath-7B (Uniform)number of samples=0.59M, Base Model=DeepSeekMath-7B, Sampling Strategy=Uniform2024.0621.3
hybridModel=LLaMA-3-8B2025.0221.3
GPT-SelectModel=Qwen2.5Math-1.5B2025.0220.6
DART-Math-Llama3-70B (Prop2Diff)number of samples=0.59M, Base Model=Llama3-70B, Sampling Strategy=Prop2Diff2024.0620
hybridModel=Qwen2.5-1.5B2025.0219.7
Llama3-70B-VRTnumber of samples=0.59M, Base Model=Llama3-70B2024.0619.3
DART-Math-Llama3-70B (Uniform)number of samples=0.59M, Base Model=Llama3-70B, Sampling Strategy=Uniform2024.0619.1
DeepSeekMath-7B-VRTnumber of samples=0.59M, Base Model=DeepSeekMath-7B2024.0619.1
ensembleModel=Qwen2.5-1.5B2025.0218.8
TATAModel=Qwen2.5-1.5B2025.0218.8
GPT-SelectModel=Qwen2.5-3B2025.0217.3
Llama2-70B-Xwin-Math-V1.1+number of samples=1.4M, Base Model=Llama2-70B2024.0616.3
Llama3-70B-MMIQCnumber of samples=2.3M, Base Model=Llama3-70B2024.0615.3
ensembleModel=LLaMA-3-8B2025.0215.3
GPT-SelectModel=Qwen2.5-1.5B2025.0214.8
DART-Math-Mistral-7B (Prop2Diff)number of samples=0.59M, Base Model=Mistral-7B, Sampling Strategy=Prop2Diff2024.0614.7
DART-Math-Llama3-8B (Prop2Diff)number of samples=0.59M, Base Model=Llama3-8B, Sampling Strategy=Prop2Diff2024.0614.5
DeepSeekMath-7B-Instructnumber of samples=0.78M, Base Model=DeepSeekMath-7B2024.0614.2
GPT-SelectModel=LLaMA-3-8B2025.0213.9
DART-Math-Llama3-8B (Uniform)number of samples=0.59M, Base Model=Llama3-8B, Sampling Strategy=Uniform2024.0613.6
DART-Math-Mistral-7B (Uniform)number of samples=0.59M, Base Model=Mistral-7B, Sampling Strategy=Uniform2024.0613.2
DeepSeekMath-7B-MMIQCnumber of samples=2.3M, Base Model=DeepSeekMath-7B2024.0613
Llama3-70B-MetaMathnumber of samples=0.40M, Base Model=Llama3-70B2024.0611.6
Llama3-70B-ICLBase Model=Llama3-70B2024.0610.8
Llama2-7B-Xwin-Math-V1.1+number of samples=1.4M, Base Model=Llama2-7B2024.0610.5
Llama3-8B-MMIQCnumber of samples=2.3M, Base Model=Llama3-8B2024.069.6
Mistral-7B-MMIQCnumber of samples=2.3M, Base Model=Mistral-7B2024.069.4
DeepSeekMath-7B-ICLBase Model=DeepSeekMath-7B2024.069.3
Llama3-8B-VRTnumber of samples=0.59M, Base Model=Llama3-8B2024.069.3
Mistral-7B-VRTnumber of samples=0.59M, Base Model=Mistral-7B2024.068.7
TATAModel=Qwen2.5-0.5B2025.028.6
hybridModel=Qwen2.5-0.5B2025.027.9
Mistral-7B-WizardMath-V1.1 (RL)Base Model=Mistral-7B2024.067.7
ensembleModel=Qwen2.5-0.5B2025.027.7
GPT-SelectModel=Qwen2.5-0.5B2025.027.1
Mistral-7B-MetaMathnumber of samples=0.40M, Base Model=Mistral-7B2024.065.9
Llama3-8B-MetaMathnumber of samples=0.40M, Base Model=Llama3-8B2024.065.5
Llama3-8B-ICLBase Model=Llama3-8B2024.064.2
Mistral-7B-ICLBase Model=Mistral-7B2024.063.7