List your company

Mathematical Reasoning on GSM8K (test) (Accuracy, Average response length)

95.7AccuracyRPAM
92.89293.62194.3595.079Jan 7, 2026

Evaluation results

MethodLinks
RPAMMethod Category=Data-dependent Merging2026.0195.71,086
CoDMethod Category=Prompt-guided2026.0195.61,017
Average MergingMethod Category=Data-free Merging2026.0195.41,088
Ada-R1Method Category=Training-based2026.0195.31,161
Qwen3-4B-ThinkingMethod Category=Base Model, Thinking Strategy=Long-CoT2026.0195.21,521
Task ArithmeticMethod Category=Data-free Merging2026.0195.21,181
DARE-LinearMethod Category=Data-free Merging2026.0195.11,641
AIMMethod Category=Data-dependent Merging2026.0195.11,060
TIES MergingMethod Category=Data-free Merging2026.0194.81,170
ACMMethod Category=Data-dependent Merging2026.0194.7925
Qwen3-4B-InstructMethod Category=Base Model, Thinking Strategy=Short-CoT2026.0193374