Reasoning Shapes Alignment: Investigating Cultural Alignment in Large Reasoning Models with Cultural Norms