Comment by storus

BM25 is quite bad and needs to be retrained for each corpus anew. SPLADEv2 is much better and there are even better sparse embeddings these days.