Pramana: Fine-Tuning Large Language Models for Epistemic Reasoning through Navya-Nyaya Paper โข 2604.04937 โข Published Feb 14 โข 1