[Enhancement]: about sparse embedding, may upper_bound is calc with max_value * query_value on this dim, instead only max_value on this dim?

[ldak4747]

Is there an existing issue for this?

• [X] I have searched the existing issues

What would you like to be added?

in src/index/sparse/sparse_inverted_index.h InvertedIndex::search_wand, when calc the upper_bound by traverse cursors, codes as "upper_bound += cursors[pivot]->max_score();", may the codes could be "upper_bound += (cursors[pivot]->max_score() * query_value);" (pseudocode)？

that maybe query value on one dim is very large, so upper_bound may not precise perfectly?

Why is this needed?

No response

Anything else?

No response

[ldak4747]

for example, index_value on one dim is from 0 to 1, but query_value on one dim is 1e10, so the inner product maybe very huge on this dim, the distance between 0.11e10 and 0.91e10 maybe very huge, but distance between 0.1 and 0.9 maybe no influence to prevent found_pivot = false

[zhengbuqian]

Hi @ldak4747, I didn't get your point. Could you elaborate further?

query_value is multiplied onto max_score when the cursor is created.