Published result

An unstated-requirement pass in Kogen shaping lost to unpatched Kogen over two Go repetitions and is dropped

A plain-language summary comes first, followed by the full text of the result's reproduction pack.

Result summary and pack text · reproduce/p2-unstated-go-pooled/README.md

In plain terms

The patch that adds an unstated-requirement pass to Kogen's shaping step was compared with unpatched Kogen over two repetitions of the same 13 Go tasks (26 pairs), same commit and model (gpt-6.1-terra, medium effort). The hidden checks passed on 19 pairs without the patch and 17 with it: one pair gained and three lost, against a keep rule of at least three net wins, and the patched build ended build_blocked 4 times against 1. The patch is dropped. This replaces the first-repetition page.

Verdict key: Yes, The result supports the change or claim that was tested. No, The result does not support it, so nothing changes. Unresolved, The evidence is not strong enough either way. Superseded, A later test replaced this result.

Full text of the reproduction pack

The text below is the pack's README as published in the repository (a system folder name may show without its leading slash, so that the site's privacy scan passes).

A Kogen patch that adds an unstated-requirement pass to shaping did not earn its keep over two Go repetitions. Across 26 task pairs (13 Go tasks, two repetitions) unpatched Kogen passed the hidden checks on 19 and the patched build on 17: one pair gained, three lost, net -2, against a keep rule of at least +3. Verdict: no: the patch is dropped. This replaces the first-repetition result p2-unstated-go-screen (unresolved). Finished 31 October 2026.

What was run

Same Kogen commit 3e759e88 (vendored Loop pin 8a2fcdae) in both arms, model gpt-6.1-terra at medium effort for shaping, building and review, request-only tasks, one trial per arm per task per repetition, cells interleaved, one Linux laptop-class host. Repetition 1 is the run behind p2-unstated-go-screen; repetition 2 is a second run of the same patch and tasks.

Arm What differs Patch sha256
b0 nothing (empty patch) none
p2 unstated-requirement pass in shaping (patch/p2-unstated.diff) aa4acc9f8325b85724699d54a36fa2f7ffba879472dddf514d9b6866d6e06c92

Result

Arm Hidden pass Wall s Tokens Requests
b0 19/26 6473 6832415 416
p2 17/26 6113 6347677 400

Wall, tokens and requests include cells that ended refused, build_blocked or review_blocked. Pairs that differ: G01r2 and G07r2 and G10r1 pass unpatched and end build_blocked with the patch; G05r2 fails unpatched (review_blocked) and passes with the patch. Outcomes: unpatched landed 23, build_blocked 1, refused 1, review_blocked 1; patched landed 20, build_blocked 4, refused 2. The patched arm ended build_blocked more often (4 against 1), which is the correctness regression that closes the question.

Limits

  • One trial per arm per pair; repetition 1 per-cell wall and token sums here are recomputed from the cell results and differ slightly from the figures printed on the first-repetition page.
  • Intent validity and acceptance separation are not scored; the numbers are hidden-check results only. The model has no listed price, so cost is token counts.
  • Cold-cache laptop-class host, so wall times do not compare with other hosts.

Reproduce the numbers

python3 -I reproduce/p2-unstated-go-pooled/report.py reproduce/p2-unstated-go-pooled/data p2-unstated-go-pooled p2 | diff - reproduce/p2-unstated-go-pooled/expected-output.txt

report.py uses only the standard library. data/p2-unstated-go-pooled_cells.tsv has one row per cell (G01 to G13 per repetition, suffix r1 or r2; names and hidden suites are not included).

Rerun

The table reruns from data/. To run the cells again you need Kogen at 3e759e88 with patch/p2-unstated.diff applied, the same 13 Go tasks and a ChatGPT login.