Repository navigation
TIERING #83: offline context tiering evaluation and go/no-go - #22
Merged
Merged
Conversation
Card 83 requires the hypotheses and evaluation corpus to be fixed before any result exists, so the comparison cannot be tuned to the outcome.
Card 83 needs a reproducible quality and cost comparison with a go/no-go. The harness reuses computeRetrievalEvalMetrics and a deterministic local embedder so the comparison runs without paid models or live Firestore, and the test pins the corpus hash and decisions.
The roadmap is the canonical record, so the decision, commits and the separation between local checks and absent deployed evidence belong there.
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Card 83. Hypotheses and a frozen synthetic corpus were committed first (fe05c41), then an offline harness compared tiering variants against the retained baseline (fadddef).
Decision: L0 first-sentence index no-go (recall@5 1.000 to 0.722). Summary and dual index no-go for now (gains rest on hand-written summaries, recall at ceiling). Generated summary as search payload is a provisional go (answerability 0.444 to 1.000, payload tokens down 55%), pending a confirmation run with real generated summaries. Nothing ships.
Docs: docs/research/2026-10-30-tiering-evaluation-plan.md and -results.md. Reproduce: npm --prefix functions run eval:tiering. Existing retrieval evaluation unchanged.
Local gates: metacortex_typecheck, metacortex_test (181 passed, 1 skipped), metacortex_package (129). No paid model, Firestore or deployment used.
🤖 Generated with Claude Code