The consonantal representation of the lexeme of a word occurrence in Hebrew script.
This feature is present on objects of type word and lex. (Only in version 2021 it is present on lex nodes as well).
Only the consonants of the lexeme are present: no vowel pointing and no other diacritical marks.
It is difficult to enter Hebrew text. One of the handiest ways to get Hebrew text into a query is to copy and paste it from some other Hebrew text, e.g. from the SHEBANQ.
You could try to copy and paste Hebrew text into a word processor, then remove all diacritics, and paste the result into your query as a value for lex_utf8.