arborist/bench
russell@unturf.com 6ba833878d
bench: progressive-AND fixture + 2026-05-09 A/B baseline report
The bench-qa-smoke A/B against commit 416f956 surfaced a methodology
gap: the smoke fixture is structurally insensitive to retrieval-side
changes — every smoke query succeeds at full-AND chain 0 on every
shard, so cache_keys match byte-for-byte BEFORE/AFTER and any
STRICT-rate variance is pure LLM dice. Discovered when 5pp signal
floor at n=3 produced ±20pp swings with identical cache_keys.

This commit pins a fixture that ACTUALLY exercises the new code path:

- 5 PROGRESSIVE queries (full-AND fails on at least one shard;
  _progressive_and_token_chains drops shortest-tokens-first). Includes
  the canonical Gundremmingen case from commit 2b9d1f0.
- 2 DF_FILTER queries (every AND chain fails on at least one shard;
  search lands in OR-mode where _filter_or_pool_by_df trims
  high-DF tokens).
- 2 CONTROLs (full-AND succeeds on every shard; cache_keys must
  remain identical BEFORE/AFTER for any retrieval-side change).

Which-chain-fires-on-which-shard was empirically verified at HEAD
against ~/.arborist/shards (4 wiki shards, 1.5M chunks each). Header
documents the chain that wins per query; re-probe if shards drift.

Baseline A/B (commit 416f956 ON vs OFF, n=3, --burn between samples,
fixture run via make bench-qa-progressive-and):

  category       cache_key parity    STRICT-rate delta    median latency delta
  PROGRESSIVE    9/15 drifted        +0.0pp (11/45 both)   -3.6s
  DF_FILTER      3/6 drifted         +5.6pp (10/18 → 11/18) -3.9s
  CONTROL        0/6 drifted         +0.0pp (15/18 both)    +0.0s

The CONTROL +0.0s latency delta is the load-bearing diagnostic —
when retrieval is byte-identical the search-side change cannot
affect total wall, and it doesn't. PROGRESSIVE / DF_FILTER cache_key
drift is the change actively redirecting retrieval, with the
expected latency win and no STRICT regression.

Future retrieval-side work (#000039 sqlite-vec backend, synonym
expansion changes, embedding rerank, hybrid scoring) should bench
against this fixture in addition to the full bench-qa scoreboard.

Hygiene:
- make test → 1623 passed, 28 skipped
- make chain-check-shards → 0 breaks across all 7 shards
2026-05-10 06:35:18 -04:00
..
batteries ticket #000032: combinatorics@v1 π* (pure-integer counting kernel) 2026-05-09 13:20:29 -04:00
fixtures textbook ingest: 4 base-knowledge additions for pillars I/II/III/IX 2026-05-09 20:38:32 -04:00
results bench: progressive-AND fixture + 2026-05-09 A/B baseline report 2026-05-10 06:35:18 -04:00
scripts textbooks: TeX-source ingest closes pillars I + IV (Hilbert + Boole) 2026-05-09 16:06:53 -04:00
emergent_log.jsonl ticket(#000006): 300-cycle update — zero false-positive STRICT post-hardening 2026-05-03 23:16:45 -04:00
qa_questions.txt aborist/arborist 2026-05-07 09:31:49 -04:00
qa_questions_metacog_subset.txt qa(#000011 + 4 more): SOFT_PREFLIGHT_HINT impl + 5-task fan-out 2026-05-03 23:00:56 -04:00
qa_questions_progressive_and.txt bench: progressive-AND fixture + 2026-05-09 A/B baseline report 2026-05-10 06:35:18 -04:00
qa_questions_quantifier_baseline.txt bench(#000008): harness extension — FC rate, violation kinds, raw brackets 2026-05-02 18:35:08 -04:00
qa_questions_quantifier_subset.txt ticket(#000008): §12 dry-run bench findings + --policy harness flag 2026-05-03 08:39:20 -04:00
qa_questions_smoke.txt speed: pytest-xdist, bench smoke, concurrency default; UTF surrogate fix 2026-05-02 09:29:40 -04:00
qa_sweep.py aborist/arborist 2026-05-07 09:31:49 -04:00
run.sh aborist/arborist 2026-05-07 09:31:49 -04:00