Most important words for false positives / false negatives

[hanshupe007]

I wonder if there is a feature to quickly see in a summary the most important words that contribute most to the false positives / false negatives. It would be a simple count and would make it easy to spot if a model did learn something usesless.

I saw that the last commit in LIT is already a while ago, is the project still maintained and continues?

Thanks!