arxiv preprint - Time Travel in LLMs: Tracing Data Contamination in Large Language Models | AI Breakdown | Podwise