From 372bebd0a376944c1679c48649f20ea42bafcc9b Mon Sep 17 00:00:00 2001 From: Bruno Postle Date: Thu, 23 Jul 2026 22:39:08 +0100 Subject: [PATCH] =?UTF-8?q?DESIGN.md:=20backfill=20=C2=A719=20with=20homem?= =?UTF-8?q?aker-py-161's=20in-search=20A/B=20verdict?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 161 (2026-07-22) answered the "remaining open question" §19 left dangling — threading fit into driver.search so shape_rotate/deslim can fire mid-GA — but the result only ever landed in bd notes, never here. Also negative: full-budget harbor-house A/B (seeds 0-3) shows no improvement, confirming the finish-time finding at in-search scale. Both halves of §19's mechanism space are now closed negative in the doc, matching bd state. Co-Authored-By: Claude Sonnet 5 Claude-Session: https://claude.ai/code/session_01GDZjAATDWW1xFfc7xnJqSt --- DESIGN.md | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/DESIGN.md b/DESIGN.md index 5273bf3..0af6654 100644 --- a/DESIGN.md +++ b/DESIGN.md @@ -2562,6 +2562,21 @@ with/without) is the remaining open question and would need to be its own measur before further code changes — this session's finding is that the *finish-time* half of the issue's candidate mechanisms is a dead end, not that geometry repair is impossible in general. +**In-search follow-up (measured, 2026-07-22, `homemaker-py-161`) — also negative.** +`driver.search`/`search_staged` gained `enable_shape_repair: bool = False`, threading a cached +`Fitness` instance into `operators.mutate()` only when set (mirrors `enable_reassociate`'s clean- +toggle pattern; default off reproduces prior runs byte-for-byte). Full A/B on harbor-house +`init.dom` cold-start, budget=1,000,000, pop=16, child_budget=80, workers=4, seeds 0–3: fails +`[14,15,12,17]` mean 14.50 (off) vs `[17,14,16,12]` mean 14.75 (on) — no improvement, and the 0.25 +delta is far inside the 12–17 seed-to-seed spread in both arms. A smaller pilot (budget=20000, 3 +seeds) matched: off mean 31.33, on mean 32.00. In-search selection pressure and population +diversity do **not** rescue `shape_rotate`/`deslim` on harbor-house-scale programmes either — the +residual fails look like a genuine floor for this representation on this programme, not an +inefficiency reachable by richer local operators, in either regime. Code kept (not reverted) for +reuse/reproducibility per the `enable_reassociate` precedent; test +`test_enable_shape_repair_threads_fit_into_mutate` in `tests/test_driver.py`. Both halves of §19's +candidate mechanism space (finish-time and in-search) are now closed negative. + ## 20. In-search global collapse (`homemaker-py-qpk`) — DONE (positive, size-dependent) **Motivation.** §17 (`94g`) landed the FINISH-TIME global cell↔room collapse — a one-shot label