Update README.md

README.md

@@ -65,8 +65,9 @@ The model adopts an MoE language model, a native-resolution visual encoder (Moon
 
 > [!Note]
 > Recommended parameter settings:
-> - For **Thinking models**, it is recommended to use `Temperature = 0.6`. 
-> - For **Instruct models**, it is recommended to use `Temperature = 0.2`. 
+> - For **Thinking models**, it is recommended to use `Temperature = 0.8`. 
+> - For **Instruct models**, it is recommended to use `Temperature = 0.2`.
+> - Greedy sampling (`Temperature = 0.0`) is okay for non-thinking (instruct) models (aligned with our evaluation setting).
 
 ## Performance