Issue: LevenshteinSimilarityIndex performs brute‑force O(N) search even with max distance – unusable for…
Issue: LevenshteinSimilarityIndex performs brute‑force O(N) search even with max distance – unusable for…: a task in LegoFlow-SWE (Harbor dataset). The LevenshteinSimilarityIndex in gensim’s similarities module exposes a max distance parameter that is documented as enabling a more efficient…
The task
The `LevenshteinSimilarityIndex` in gensim’s similarities module exposes a `max_distance` parameter that is documented as enabling a more efficient search. However, the current implementation always scans the entire vocabulary linearly for every query, making it O(N) regardless of `max_distance`. The parameter is…
Part of Lego-X/LegoFlow-SWE.