Update README.md

README.md

@@ -108,7 +108,7 @@ This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing
 | Correct predictions   | 518                            | 394                            | 516                            |
 | Accuracy             | 10.19%                         | 7.75%                          | 10.15%                         |
 | Most common move     | d (41.14%)                     | d (67.61%)                     | a (38.72%)                     |
-| Middle column usage  | 75.05%                         | 99.53%                         | 27.38%                         |
+| Middle column usage  | 75.05%                         | 99.53%                         | 29.08%                         |
 
 *(Middle column usage = c (17.34%) + d (2.70%) + e (9.04%) = 27.38%)*