- Maximum Entropy Models handle overlapping features well
- unlike Naive Bayes Model, there is no double counting
- Maximum Entropy Models handle overlapping features well, but do not automatically model feature interactions. They do not model interaction terms, you have to add them