Imagine having an AI assistant that could accurately answer any health question you have, providing reliable information backed by scientific literature. That's the vision behind the BioGen track at TREC 2024, a leading information retrieval conference. This year, researchers tackled the critical challenge of ensuring that Large Language Models (LLMs) don't just generate convincing-sounding answers, but also ground those answers in verifiable medical evidence. Why is this so important? Because in the health domain, misinformation can have serious consequences. Current LLMs, while impressive in their language abilities, are prone to 'hallucinations'—generating incorrect or misleading information. The BioGen track focused on 'reference attribution,' essentially teaching LLMs to cite the scientific papers supporting their claims. Teams participating in the challenge employed various techniques, primarily using a two-stage approach. First, they used advanced search methods to retrieve relevant documents from a massive database of medical literature. Then, they prompted LLMs to generate answers based on these retrieved documents, including proper citations (PMIDs). The results are promising, with some systems demonstrating high accuracy in answering questions and providing supporting evidence. However, challenges remain, particularly in ensuring the relevance and correctness of the cited references. This research lays the groundwork for developing more trustworthy AI systems for biomedical information access. Imagine a future where patients and clinicians can confidently rely on AI to provide accurate, evidence-based answers to complex medical questions – the BioGen track is a crucial step in making this vision a reality.