Randomness, Not Representation: Evaluating Cultural Alignment in LLMs - View it on GitHub