Hi team! Could you please help me to find a parameter, which controls this behaviour? I have a dataset with one string facet with values

facet1

,

facet2

. If I search “foo bar” without chosen facets, it may return to me document

{"facet": "facet1", "text": "foo ba"}

and 0 in facet_count for

facet2

. But if I choose facet2 instead, it will show me more extreme typos, like

fow bas

. So it looks like typesense when nothing is found switching to more tolerant rules to try find anything. I checked and

exhaustive_search

is default in my case