Formal models and scripts: cosine-zero-query-vector-silent-empty-results-known-15540-row-or-dnp

Tools: Python

These are the working models and scripts this part of the hunt used. Some encode the code as it is (a counterexample is the expected result), others a fixed or correct variant; read the header comment of each file. Paths inside them were replaced by placeholders such as $LUCENE (a Lucene checkout) and $WORK.

Findings

• kNN and vector-similarity queries accept an all-zero query target on a COSINE field, which the field itself rejects; the cosine is NaN, so the search returns no hits or the HNSW entry document with score NaN, with no error (finding LFV-1141; repro)

Files

• model/cosine_zero_target.py (used for LFV-1141)

How to run

• Python: python3 <file>.py; scripts that import z3 need the z3-solver package.

Snapshot of 2026-11-04. AI-assisted; not yet triaged by a human Lucene committer. Feedback: open an issue in lucene-ai-bug-hunt.