Gensim lda gives negative log-perplexity value - is it normal and how can i interpret it?

Question

I am currently using Gensim LDA for topic modeling.

While Tuning hyper-parameters I found out that the model always gives negative log-perplexity

Is it normal for model to behave like this?? (is it even possible?)

if it is, is smaller perplexity better than bigger one? (-100 is better than -20??)

Answer 1

It is usual for gensim to give a negative value, and here, LDA -20 perplexity is better, as you should see the value closer to 0.