List your company

Mathematical Problem Solving on Omni-MATH (Full Set)

64AccuracyCIKA
24.4834.744555.26May 8, 2026

Evaluation results

MethodLinks
CIKACausal=true, Model Scale (>7B)=false2026.0564
CIKACausal intervention=true2026.0564
o1-miniCausal=false, Model Scale (>7B)=true2026.0560.5
o1-miniCausal intervention=false2026.0560.5
DSR1Causal=false, Model Scale (>7B)=true2026.0555
DeepSeek-R1Causal intervention=false2026.0555
Qwen-72BCausal=false, Model Scale (>7B)=true2026.0540
SRV (base)Causal=false, Model Scale (>7B)=false2026.0526