Top2Vec
Top2Vec copied to clipboard
Contextual document spans
In contextual top2vec there are multiple document vectors per document. Is there a way to recover which documents these vectors represent? Or what document span (snippet of text) corresponds to each document vector?