Update README.md

README.md

@@ -36,7 +36,7 @@ On SWE-Bench Verified, **KAT-Dev** achieves comparable performance with **62.4%*
   <tbody>
     <tr>
       <td><strong>1. Mid-Training</strong></td>
-      <td>We observe that adding extensive training for tool-use capability, multi-turn interaction, and instruction-following at this stage may not yield large performance gains in the current results (e.g., on leaderboards like SWE-bench), but it will have a significant impact on the subsequent SFT and RL stages. This suggests that improving these foundational capabilities can have a profound effect on the model’s ability to perform complex tasks.
+      <td>We observe that adding extensive training for tool-use capability, multi-turn interaction, and instruction-following at this stage may not yield large performance gains in the current results (e.g., on leaderboards like SWE-bench). However, since our experiments are based on the Qwen3-32B model, we find that enhancing these foundational capabilities will have a significant impact on the subsequent SFT and RL stages. This suggests that improving such core abilities can profoundly influence the model’s capacity to handle more complex tasks.
 </td>
     </tr>
     <tr>