I'm working on a Text classification Problem with Keras and lime. I'm using neural networks and word embeddings.
Afterwards I want to explain my model with lime.
When I use lime to explain the first three sentences of the test data lime predicts for me the following labels: 1 0 1 1 0 1. But my Keras model predicts: 1 0 1 1 1 1
Also the probabilities for label 1 or label 0 differ for the Keras model and the lime output.
Why is this the case or is there something wrong with my model?
I'm working on a Text classification Problem with Keras and lime. I'm using neural networks and word embeddings.
Afterwards I want to explain my model with lime. When I use lime to explain the first three sentences of the test data lime predicts for me the following labels: 1 0 1 1 0 1. But my Keras model predicts: 1 0 1 1 1 1 Also the probabilities for label 1 or label 0 differ for the Keras model and the lime output.
Why is this the case or is there something wrong with my model?
Thank you in advance!