ResearchBenchmarksMathematical Problem Solving on MathArena FA CMIMC 2025Follow13Accuracy (Ratio)CIKA12.3512.6751313.325May 8, 2026Evaluation ResultsMethodMethodLinksAccuracy (Ratio)AccuracyCIKAsampling=pass@1, shots...sampling=pass@1, shots=1-shot2026.051332.5