Fathom-Vaidya: Advancing Medical Reasoning with Rubric-Based Rewards.

arXiv:2609.24480v1 Announce Type: new Abstract: Deploying Large Language Models (LLMs) in healthcare requires robust performance across two complementary dimensions - diagnostic reasoning: the convergent, evidence-driven task of inferring a patient's condition