List your company

Mathematical Reasoning on SVAMP (Accuracy (%))

95.2Accuracy (%)MetaEvo
-2.071223.181948.43573.6881May 29, 2026Jun 3, 2026Jun 8, 2026Jun 14, 2026Jun 19, 2026Jun 24, 2026Jun 30, 2026

Evaluation results

MethodLinks
MetaEvoBackbone=Qwen2.5-14B-Instruct2026.0595.2
MetaEvoBackbone=LLaMA3.1-8B-Instruct2026.0593.8
MetaEvo - w/o MOBackbone=Qwen2.5-14B-Instruct2026.0593.8
MetaEvo - w/o CDABackbone=Qwen2.5-14B-Instruct2026.0593.7
Self-DiscoverBackbone=Qwen2.5-14B-Instruct2026.0593.4
SE-GPTBackbone=Qwen2.5-14B-Instruct2026.0593.3
Self-RefineBackbone=Qwen2.5-14B-Instruct2026.0591.8
Base ModelBackbone=Qwen2.5-14B-Instruct2026.0591.5
Self-ICLBackbone=Qwen2.5-14B-Instruct2026.0590.3
MetaEvo - w/o MOBackbone=LLaMA3.1-8B-Instruct2026.0590.1
Self-DiscoverBackbone=LLaMA3.1-8B-Instruct2026.0588.7
SE-GPTBackbone=LLaMA3.1-8B-Instruct2026.0587.1
Hard-Routed MoR-LoRABackbone=LLaMA-8B2026.0685.67
IDBackbone=Qwen2.5-3B, Fine-tuning Dataset=Meta-MathQA, Subset Fraction=50%2026.0685.4
Self-ICLBackbone=LLaMA3.1-8B-Instruct2026.0585
IWDBackbone=Qwen2.5-3B, Fine-tuning Dataset=Meta-MathQA, Subset Fraction=50%2026.0685
GREATSBackbone=Qwen2.5-3B, Fine-tuning Dataset=Meta-MathQA, Subset Fraction=50%2026.0684.8
COLMBackbone=Qwen2.5-3B, Fine-tuning Dataset=Meta-MathQA, Subset Fraction=50%2026.0684.2
GradNormBackbone=Qwen2.5-3B, Fine-tuning Dataset=Meta-MathQA, Subset Fraction=50%2026.0684.1
PartitionSelBackbone=Qwen2.5-3B, Fine-tuning Dataset=Meta-MathQA, Subset Fraction=50%2026.0684.1
LoRiModel=LLaMA-3.1-8B, Explicit CoT=false2026.0683.3
LoRi-EModel=LLaMA-3.1-8B, Explicit CoT=false2026.0683.3
Base ModelBackbone=LLaMA3.1-8B-Instruct2026.0583.3
Self-RefineBackbone=LLaMA3.1-8B-Instruct2026.0583.2
RandomBackbone=Qwen2.5-3B, Fine-tuning Dataset=Meta-MathQA, Subset Fraction=50%2026.0683.2
BaselineBackbone=LLaMA-8B2026.0682
Hard-Routed MoR-LoRABackbone=LLaMA-3B2026.0681.33
MetaEvo - w/o CDABackbone=LLaMA3.1-8B-Instruct2026.0581.3
SFT-CoTModel=LLaMA-3.1-8B, Explicit CoT=true2026.0681
SIM-CoTModel=LLaMA-3.1-8B, Explicit CoT=false2026.0679.4
LoRi-EModel=LLaMA-3.2-3B, Explicit CoT=false2026.0679.3
SFT-CoTModel=LLaMA-3.2-3B, Explicit CoT=true2026.0678.7
CODIModel=LLaMA-3.1-8B, Explicit CoT=false2026.0678.1
LoRiModel=LLaMA-3.2-3B, Explicit CoT=false2026.0675
SIM-CoTModel=LLaMA-3.2-3B, Explicit CoT=false2026.0674.9
KAVAModel=LLaMA-3.2-3B, Explicit CoT=false2026.0672.7
CODIModel=LLaMA-3.2-3B, Explicit CoT=false2026.0672.4
PCCoTModel=LLaMA-3.2-3B, Explicit CoT=false2026.0669.5
FRAMEBackbone=LLaMA-3.1-8B2026.0668.4
MoABackbone=LLaMA-3.1-8B2026.0667.3
FourierMoEBackbone=LLaMA-3.1-8B2026.0667.1
FlyLoRABackbone=LLaMA-3.1-8B2026.0667
HMoRABackbone=LLaMA-3.1-8B2026.0666.7
MixLoRABackbone=LLaMA-3.1-8B2026.0666.3
HydraLoRABackbone=LLaMA-3.1-8B2026.0665.8
DoRABackbone=LLaMA-3.1-8B2026.0665.2
FourierFTBackbone=LLaMA-3.1-8B2026.0664.7
LoRABackbone=LLaMA-3.1-8B2026.0664.5
LoRi-EModel=LLaMA-3.2-1B, Explicit CoT=false2026.0664.3
LoRiModel=LLaMA-3.2-1B, Explicit CoT=false2026.0664
LoRiModel=Qwen2.5-0.5B, Explicit CoT=false2026.0663
LoRi-EModel=Qwen2.5-0.5B, Explicit CoT=false2026.0662.7
SFT-CoTModel=Qwen2.5-0.5B, Explicit CoT=true2026.0662.3
CODIModel=LLaMA-3.2-1B, Explicit CoT=false2026.0661.1
SFT-CoTModel=LLaMA-3.2-1B, Explicit CoT=true2026.0661
COLMBackbone=Llama-3.1-8B, Training Dataset=MetaMathQA, Subset Ratio=12.5%2026.0660.8
KAVAModel=LLaMA-3.2-1B, Explicit CoT=false2026.0658.9
IDBackbone=Llama-3.1-8B, Training Dataset=MetaMathQA, Subset Ratio=12.5%2026.0658.4
PCCoTModel=LLaMA-3.2-1B, Explicit CoT=false2026.0657.7
NoCoTModel=LLaMA-3.2-3B, Explicit CoT=false2026.0656.9
PartitionSelBackbone=Llama-3.1-8B, Training Dataset=MetaMathQA, Subset Ratio=12.5%2026.0656.7
NoCoTModel=LLaMA-3.1-8B, Explicit CoT=false2026.0655.3
GradNormBackbone=Llama-3.1-8B, Training Dataset=MetaMathQA, Subset Ratio=12.5%2026.0654.1
KAVAModel=Qwen2.5-0.5B, Explicit CoT=false2026.0650.6
COCONUTModel=LLaMA-3.2-1B, Explicit CoT=false2026.0648.8
CODIModel=Qwen2.5-0.5B, Explicit CoT=false2026.0647
NoCoTModel=LLaMA-3.2-1B, Explicit CoT=false2026.0644.1
iCoTModel=LLaMA-3.2-1B, Explicit CoT=false2026.0640.9
NoCoTModel=Qwen2.5-0.5B, Explicit CoT=false2026.0634.5
PCCOTModel=Qwen2.5-0.5B, Explicit CoT=false2026.0633
BaselineBackbone=LLaMA-3B2026.061.67