Repository navigation
Conversation
1 of 2 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Describe the changes in the pull request
NaviX revisits shared neighbors and currently repeats the NO/MAYBE callback, including document-summary lookups. Cache each HNSW node's decision across batches so the callback runs at most once per node between resets. Exact filtering remains the consumer's responsibility.
The cache allocates one byte per graph node on its first use (~1 MB for a one-million-node graph), uses the iterator allocator, grows safely for newly migrated backend nodes, and is invalidated on reset. Non-NaviX queries do not allocate it. The existing stable index/label/filter-view contract still applies. Filter check/rejection counters count callback evaluations, excluding cache hits.
Stacked on #1113. This draft contains only caching; traversal policy and sparse-query recovery are unchanged. No benchmarks were run and no throughput improvement is claimed.
Which issues this PR fixes
Follow-up to the experimental NaviX POC; does not close an issue.
Main objects this PR modified
Mark if applicable
Validation: focused Debug build and all seven
test_navixcases, including the existing timeout/reset and frontend cases. Formatting andgit diff --checkpassed. Benchmarks were intentionally not run.