Update README.md

README.md

@@ -24,8 +24,8 @@ tags:
 We release our process reward model effectively trained with active learning based on `Qwen/Qwen2.5-Math-7B-Instruct`. The model achieved 75.0% on [ProcessBench](https://github.com/QwenLM/ProcessBench) and 65.5% on [PRMBench](https://github.com/ssmisya/PRMBench).
 
  <p align="center">
-  <img src="https://raw.githubusercontent.com/sail-sg/ActivePRM/refs/heads/master/assets/processbench.png" width=70%/>
-  <img src="https://raw.githubusercontent.com/sail-sg/ActivePRM/refs/heads/master/assets/prmbench.png" width=70%/>
+  <img src="https://raw.githubusercontent.com/sail-sg/ActivePRM/refs/heads/master/assets/processbench.png" width=90%/>
+  <img src="https://raw.githubusercontent.com/sail-sg/ActivePRM/refs/heads/master/assets/prmbench.png" width=90%/>
 </p>
 
 ### Usage