edits

README.md

@@ -242,7 +242,7 @@ Muon represents a significant advancement in neural network optimization by addr
 As you build and train your own models, consider Muon for hidden layer optimization, especially during pre-training phases where building new capabilities is the priority.
 
 ---
-💡 If this saved you from math-induced tears, consider giving it a ❤️ so my repo doesn’t feel lonely.
+👉 If you find this useful, please ⭐️ the repo or cite it in your work — it helps support the project.
 
 ## 📖 Citation