Update README.md

README.md

@@ -10,7 +10,7 @@ observation -> ... -> final answer); it never answers single-shot. Three role
 personas: **Anatomist** (organ segmentation / volumetry), **Radiologist**
 (cross-visit image comparison), **Consultant** (report-grounded synthesis).
 Paper: *From Given to Gathered Evidence: Agentic Learning for Longitudinal Medical Reasoning*.
-
+arxiv.org/abs/2609.39566
 SFT-initialized agent (40% of the training split) - already tool-calling capable;
 this is the initialization that the RL models continue from.