Update README.md

README.md

@@ -27,6 +27,8 @@ It leverages OpenAI’s `text-embedding-3-large` with a multi-head classifier to
 - Self-harm (`self_harm_l1`, `self_harm_l2`)
 - Other misconduct (`all_other_misconduct_l1`, `all_other_misconduct_l2`)
 
+Further details on the benchmark results and training methodology are in our [report](https://arxiv.org/abs/2507.15339).
+
 ---
 
 # Taxonomy