Med-V1: Small Language Models for Zero-shot and Scalable Biomedical Evidence Attribution.

arXiv:2603.05308v4 Announce Type: replace-cross Abstract: Assessing whether an article supports an assertion is essential for hallucination detection and claim verification. While large language models (LLMs) have the potential to automate this task, achieving