# In an SVM, does a more negative/positive decision score mean that it is further from the seperating hyperplane?

For example, if I have a sample with a decision score of -6 and another with a score of -3, which sample is closer to the hyperplane?

Also, does the probability of a sample belonging to a class increase with its respective distance from the hyperplane (larger distance from hyperplane = larger probability)?

I am using sklearn's decision_function and predict_proba functions to get these values if it makes a difference.

Thanks.