diff --git a/examples/class_increment_semantic_segmentation/lifelong_learning_bench/README.md b/examples/class_increment_semantic_segmentation/lifelong_learning_bench/README.md
index 5944e389..5eb1181f 100644
--- a/examples/class_increment_semantic_segmentation/lifelong_learning_bench/README.md
+++ b/examples/class_increment_semantic_segmentation/lifelong_learning_bench/README.md
@@ -92,14 +92,14 @@ The final output might look like this:
 |  1   | erfnet_lifelong_learning |  0.027414088670437726 | 0.010395591126145793 | 0.002835451693721201 | lifelonglearning | BaseModel | TaskDefinitionByDomain | TaskAllocationByDomain |         0.0001          |        1         | ['Cityscapes', 'Synthia', 'Cloud-Robotics'] | ['Cityscapes', 'Synthia', 'Cloud-Robotics'] | 2023-09-26 20:13:21 | ./ianvs-workspace/mdil-ss/lifelong_learning_bench/benchmarkingjob/erfnet_lifelong_learning/3a8c73ba-5c64-11ee-8ebd-b07b25dd6922 |
 
 
-In addition, in the log displayed at the end of the test, you can see the accuracy of known and unknown tasks in each round, as shown in the table below.
+In addition, in the log displayed at the end of the test, you can see the accuracy of known and unknown tasks in each round, as shown in the table below (in the testing phase of round 3, all classes are seen).
 
 
-| Round | Unseen Class Accuracy | Seen Class Accuracy |
+| Round | Seen Class Accuracy | Unseen Class Accuracy |
 |:-----:|:---------------------:|:-------------------:|
-|   1   |        0.0276         |       0.0293        |
-|   2   |        0.0316         |       0.0265        |
-|   3   |        0.0000         |       0.0282        |
+|   1   |        0.176         |       0.0293        |
+|   2   |        0.203         |       0.0265        |
+|   3   |        0.311         |       0.0000        |