benchmarks: shipped 8M default measured end-to-end — 84.0% hit@5, not 74% (SHA-323) - #47
Merged
Merged
Conversation
… 74% (SHA-323) The default-model row was the in-harness scan without the MaxSim rerank. Measured through the real search tool (seekstone#305, re-run 2026-09-04): 84.0% overall / 85.0% held-out / 88% on description-style queries at ~35 ms warm p50 — the latency correction is honest in the other direction (the old ~15 ms was also scan-only). Competitor rows refreshed from the same re-run so the 'same run' claim stays true; gate-v2 verdict unchanged.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
This branch was successfully deployed
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Mirrors seekstone#305: the retrieval runner never measured the 8M default through the real search tool, so the published row was the in-harness scan without the MaxSim rerank. Re-measured end-to-end 2026-09-04, same-run head-to-head refreshed via local Ollama.
Changes on /benchmarks
faq.ts: semantic latency ~14 ms → ~35 ms with the MaxSim rerank.Every competitor quality number reproduced exactly; gate-v2 verdict unchanged (fail — published miss stands). Site builds clean.