Skip to content

fix(complexity,#17151): 4 findings Complexity -- cross-links RL, stale-claims, banniere, preface 06b - #19781

Merged
myia-ai-01 merged 1 commit into
mainfrom
fix/17151-complexity-findings
Oct 8, 2026
Merged

myia-ai-01 merged 1 commit into
mainfrom
fix/17151-complexity-findings

Conversation

@jsboige

@jsboige jsboige commented Oct 7, 2026 •

Copy link
Copy Markdown
Owner

Grain: MED/notebook-python -- lane myia-po-2026:CoursIA-2 -- prev: DEEP/notebook-python #19773

fix(complexity,#17151): 4 findings Complexity -- cross-links RL, stale-claims, banniere, preface 06b

Issue : #17151 (audit partition Hermes campagne #17073) -- tranche 2/4 du dispatch ai-01 tapis central 06/10 (file de 4 issues : ordre/interp, Complexity findings, READMEs, densite).

4 findings dispatches, 4 corriges (squash c.1422)

4 corrections CONFIRMED firsthand (G.9 verifie) :

Squash c.1422 -- fix kernel drift guard

A c.1422 le gate PR gate rendait FAIL -- failing checks: Kernel drift guard (base vs PR) parce que actions/checkout@v4 (fetch-depth=1 par defaut) ne lisait pas le commit 81c32feb18 (1er des 2 commits de la PR), donc git merge-base origin/main HEAD echouait avec error: Could not read 81c32feb18....

Le fix est un squash des 2 commits initiaux en 1 seul (c692a3e17e) pousse avec --force-with-lease (branche de PR a lane unique, tel autorise par git-workflow.md) : le runner voit alors le commit de tete comme seul commit, fetch-depth=1 le couvre, et le merge-base fonctionne. Meme delta (4 fichiers, 78 insertions, 9 suppressions), seul le SHA de tete change. Le DWELL se re-arme (commit de code ordinaire, cf git-workflow.md), delai 120 min avant eligibilite merge ; le kernel drift guard devrait passer au prochain rollup.

Conformite

Perimetre

  • 4 fichiers, 78 insertions, 9 suppressions.
  • 0 secret, 0 binaire, 0 fichier hors scope.
  • Branche : fix/17151-complexity-findings (depuis origin/main).
  • 1 commit (squash depuis 2 commits initiaux, fix kernel drift guard) : c692a3e17e.

Suite de la file dispatch

Le dispatch ai-01 06/10 a 4 issues : #18053 (done c.1418, #19773), #17151 (cette PR, 4/4 complete), #18064 (READMEs, tranche 3/4, PR #19783 livree c.1421 GameTheory 15g), #13410 (densite, tranche 4/4).

🤖 Generated with Claude Code

@github-actions

github-actions Bot commented Oct 7, 2026

Copy link
Copy Markdown
Contributor

Notebook outputs-required (H.4 schema): PASS (every code cell carries an outputs: list)

@github-actions github-actions Bot added the variation-tag-missing PR sans tag Grain: <TIER>/<GENRE> (variation-protocol) label Oct 7, 2026
@github-actions

github-actions Bot commented Oct 7, 2026

Copy link
Copy Markdown
Contributor

Grain tag obligatoire (#10045, bloquant).

Grain tag absent (no Grain: / in body).

Pour passer ce gate, le body doit porter en tete une ligne de la forme :

Grain: <DEEP|MED|LIGHT>/<genre> -- lane <machine:workspace> -- prev: <TIER>/<GENRE> #<PR>

Le <genre> doit figurer dans l'enumeration §1 de variation-protocol.md (lean, qc, training, genai, notebook-python, notebook-dotnet, notebook-lean, slides, docs, guard, refactor, ledger, readme, test, tooling, research-code). Les 3 formes tolerées par l'extracteur : Grain: TIER/GENRE, **Grain:** TIER/GENRE, ## Grain + tag sur la ligne suivante. La lane doit suivre le format <machine>:<workspace> (cf. lane-claim-protocol.md).

@github-actions

github-actions Bot commented Oct 7, 2026

Copy link
Copy Markdown
Contributor

⚠️ Prose/output review needed in the notebooks this PR changed: a numeric value is not anchored, an explicit relation is contradicted, or its evidence is missing. These cases remain distinct in the JSON report; the signal is advisory, NOT a merge gate.

Scope = notebooks CHANGED in this PR, not the whole corpus. Explicit claim-check relations resolve only against named CLAIM_METRICS from the local output window and are classified SUPPORTED, CONTRADICTED, or UNPROVEN.
The markdown-claims-output-report run artifact contains the structured JSON report. See python scripts/check_markdown_claims_output.py --help for re-running locally.
Detector rationale: c.290 / c.331 / PR #11435 numeric pathology, extended with low-noise relational evidence.

@github-actions

github-actions Bot commented Oct 7, 2026

Copy link
Copy Markdown
Contributor

✅ No unanchored measurement claim detected in the notebooks this PR changed.

Scope = notebooks CHANGED in this PR, not the whole corpus. The stale-claim-report run artifact holds the structured JSON.
Rationale: the sibling detector above only compares a claim to the outputs of the cells that PRECEDE it; a claim written in a cell that precedes its code (App-5-Timetabling c.2/c.4) is invisible to it, and a value imported from a twin notebook is never produced locally. See python scripts/check_stale_claims.py --help.

@github-actions

github-actions Bot commented Oct 7, 2026

Copy link
Copy Markdown
Contributor

✅ No factual mislabel detected in the notebooks this PR changed (entity counts and tuple formulas checked against nearby committed streams).

Scope = notebooks CHANGED in this PR, not the whole corpus. The factual-mislabel-report run artifact holds the structured JSON.
Rationale: pure ABSENCE of a claimed value is the sibling stale-claim detector's job; this one only reports CONTRADICTIONS between an adjacent code cell's stream and the markdown that describes it. See python scripts/check_factual_mislabel.py --help.

@github-actions

github-actions Bot commented Oct 7, 2026

Copy link
Copy Markdown
Contributor

No organ-duplication: no added def/class collides with another series organ API (scripts/audit/organ_api_index.yaml).

Detector: python scripts/audit/detect_organ_duplication.py --base <merge-base> --body-file <pr body>
Rationale: #16776 / #13564 (rule merged in #16778).

@github-actions github-actions Bot added the consecutive-code-cells Modified notebook has >=2 consecutive code cells (#12797) label Oct 7, 2026
@github-actions

github-actions Bot commented Oct 7, 2026 •

Copy link
Copy Markdown
Contributor

Notebook PR Validation: PASS

  • Notebooks checked: 4
  • Code cells validated: 44
  • Result: All passed

Checks: H.1 (no errors), H.3 (execution_count), C.1 (no banned patterns)
Non-Python kernels (.NET/Lean): C.1 + errors only (execution_count advisory)
QuantConnect notebooks: C.1 + errors only (require QC Cloud for execution)

@github-actions

github-actions Bot commented Oct 7, 2026 •

Copy link
Copy Markdown
Contributor

Golden-Set Execution (H.7 P3)

✅ 9/9 notebooks passed (certified reproducible)

Notebook Status Time
2.1-Workflow-ML.ipynb ✅ SUCCESS 3.4s
2.2-Descente-de-gradient.ipynb ✅ SUCCESS 3.5s
2.3-Regression-lineaire-logistique.ipynb ✅ SUCCESS 4.6s
2.4-Arbres-Forets-Ensembles.ipynb ✅ SUCCESS 4.5s
Search-01-StateSpace.ipynb ✅ SUCCESS 3.3s
SL-1-LogicalLearning.ipynb ✅ SUCCESS 2.1s
RL-04-Bandits-Manchots-Python.ipynb ✅ SUCCESS 17.6s
GameTheory-04c-NashExistence-Python.ipynb ✅ SUCCESS 2.8s
GameTheory-13d-Optimistic-CFR-Python.ipynb ✅ SUCCESS 10.5s

Pinned lockfile: scripts/notebook_tools/golden_set.lock.txt (H.7 P3, axe A #4208)

@jsboige jsboige changed the title fix(complexity,#17151): 3 findings Complexity -- cross-links RL cliquables, stale-claim, banniere perimee fix(complexity,#17151): 4 findings Complexity -- cross-links RL, stale-claims, banniere, preface 06b Oct 7, 2026
@jsboige

jsboige commented Oct 7, 2026

Copy link
Copy Markdown
Owner Author

[INFO] 4/4 findings corriges -- dispatch ai-01 tapis central 06/10 tranche 2/4 complete

Le 4e finding (Complexity-06b prerequisite-gap Bell/GHZ/ket) est livre dans le commit c43aaf515d : encart markdown ## 0.1. Prérequis quantiques : qubit, ket, Bell, GHZ insere entre cell 2 et cell 3 (31 -> 32 cellules, 0 re-exec, outputs pre-existants preserves).

Resume du perimetre final :

  • 4 fichiers (Complexity-04b, 05, 05b, 06b)
  • 78 insertions, 9 suppressions
  • 4 corrections CONFIRMED firsthand (G.9 verifie) : 04b cell 20 cross-links RL cliquables ; 05 cell 21 stale-claim ; 05b cell 3 stale-claim banniere ; 06b cell 3 preface pedagogique quantique.
  • H.3 PASS, C.1/C.2/C.3 OK.

Etat de la PR : OPEN, ms=BLOCKED, m=MERGEABLE. Blocage par check-suite tierce claude (app_id 1236702) en queue depuis 20:49:57Z (~1h20) -- meme pattern que #19771 et #19773 (Tell c.1398-r7-L1 ★). Si la queue se libere, la PR est ripe-surfondre ; sinon ripe-signal coord pour [OVERRIDE].

-- myia-po-2026:CoursIA-2, c.1421

@github-actions

github-actions Bot commented Oct 7, 2026

Copy link
Copy Markdown
Contributor

G-VAR-2/3 GENRE signals (advisory, non bloquant, #10020).
La lane `myia-po-2026:CoursIA-2` voit ces signaux actifs sur les mergees du jour (UTC 2026-10-07) :

G-VAR-2 plafonne a max(1, grains_mergees_du_jour // 3) LIGHT par lane et par jour, toutes categories LIGHT confondues -- un RATIO, pas un plafond plat ; le cap calcule du jour est dans le tally ci-dessus. G-VAR-3 interdit deux genres LIGHT consecutifs. Les signaux ci-dessus rendent le fait VISIBLE (labels variation-tier-inflation, `variation-genre-run`, `variation-genre-cap-exceeded`, `variation-genre-mismatch`, `variation-genre-unknown`) -- la decision de merge reste au coordinateur.

…e-claims, banniere, preface 06b

Issue : #17151 (audit partition Hermes campagne #17073) -- tranche 2/4 du
dispatch ai-01 tapis central 06/10 (file de 4 issues : ordre/interp,
Complexity findings, READMEs, densite).

4 corrections CONFIRMED firsthand (G.9 verifie) :

* **Complexity-04b cell 20 (navigation-misplaced)** : 5 cross-links en
  texte brut (Complexity-04c, 04d, RL/RL-16, GameTheory, #15944)
  transformes en liens markdown cliquables. Hermes po-2026 partition
  Hermes 28/09 04:42Z, 1 signal sur 6 organes.

* **Complexity-05 cell 21 (stale-claim)** : la table de conclusion
  disait `facteur proche de n a chaque taille`, mais la mesure est
  1.1 (n=3), 1.4 (n=4), 6.6 (n=9) -- le facteur ne rejoint $n$
  qu'au-dela de n approx 5. Hermes po-2026 03/10 23:45Z (audit
  0357e35/#18819). Correctif minimal : `facteur qui croit vers n
  entre n=5 et n=10 -- overhead domine en deça`.

* **Complexity-05b cell 3 (stale-claim banniere)** : le print
  d'imports annoncait `Complexity-03 -- Aaronson-Arkhipov (2011)`
  alors que le carnet est le 5b (BosonSampling). Renumerotation du
  24/09 (#17593, puis #17786) a laisse cette banniere orpheline des
  le commit d'ajout. Fix = `Complexity-05b -- Aaronson-Arkhipov
  (2011), la permanente comme frontiere (BosonSampling)`.

* **Complexity-06b cell 3 (prerequisite-gap Bell/GHZ/ket)** : le
  carnet introduisait la notation ket, les etats de Bell, GHZ-8,
  sans preface pedagogique sur les prerequis quantiques. Hermes
  po-2026 30/09 10:50Z (head `e8695322`, 31 cellules). Encart
  markdown (nouvelle section `## 0.1. Prérequis quantiques : qubit,
  ket, Bell, GHZ`) insere entre la section 0 et les imports. Definit
  : qubit + base computationnelle, etats |+>/|->, produit
  tensoriel 2^n, etat de Bell |Phi+> et son groupe stabilisateur,
  etat GHZ-n et son stabilisateur dominant X_0...X_{n-1} + Z_0 Z_j.
  Explicitement borne : rappel notationnel, PAS une intro QM ni un
  raccourci vers le formalisme stabilisateur (qui est construit et
  prouve dans les cellules suivantes). Encart c.1421 (initialement
  commit `c43aaf515d`, rebase squashing suite fix kernel drift guard).

## Conformite

- C.1 : pas d'erreur volontaire (markdown/text seul, 4/4 findings).
- C.2 : 0 re-execution, outputs pre-existants preserves (insertion
  d'une cellule markdown entre cell 2 et cell 3 dans 06b ; aucun
  code touche).
- C.3 : audit re-verify 4/4 = 100 % sur ce sous-echantillon, 40 %
  de precision globale documentee (cf #18053 et #17151 body).
- H.3 : pre-commit auto-fix `source-list-missing-newlines` PASS,
  notebooks gardent execution_count et outputs pre-existants.

## Perimetre

- 4 fichiers, 78 insertions, 9 suppressions.
- 0 secret, 0 binaire, 0 fichier hors scope.
- Branche : `fix/17151-complexity-findings` (depuis origin/main).
- 1 commit (squash depuis 2 commits initiaux, fix kernel drift
  guard -- actions/checkout@v4 fetch-depth=1 ne lisait pas
  81c32fe ; squash deplace tout le delta sous c43aaf5 pour
  que merge-base origin/main HEAD fonctionne sur le runner).

## Suite de la file dispatch

Le dispatch ai-01 06/10 a 4 issues : #18053 (done c.1418, #19773),
#17151 (cette PR, 4/4 complete), #18064 (READMEs, tranche 3/4),
#13410 (densite, tranche 4/4).

Co-Authored-By: Claude Haiku 4.5 (1M context) <noreply@anthropic.com>
@jsboige
jsboige force-pushed the fix/17151-complexity-findings branch from c43aaf5 to c692a3e Compare October 7, 2026 21:24
@jsboige

jsboige commented Oct 7, 2026

Copy link
Copy Markdown
Owner Author

[INFO c.1422 squash + force-push] PR #19781 -- 4/4 findings, kernel drift guard leve

A c.1422 (22:55Z) le PR gate rendait FAIL sur Kernel drift guard (base vs PR) avec error: Could not read 81c32feb18092becdbbd8a9d62f85f2b8ba5be0a + fatal: 2e89158cda...c43aaf515d...: no merge base. Cause unique : actions/checkout@v4 (fetch-depth=1) ne lit que le commit de tete, pas les commits anterieurs, donc git merge-base origin/main HEAD echouait sur le commit 81c32feb18 (1er des 2 commits initiaux de la PR).

Fix : squash des 2 commits en 1 seul (c692a3e17e), force-push avec --force-with-lease (tel autorise par git-workflow.md -- branche de PR a lane unique). Meme delta (4 fichiers, 78/9), seul le SHA de tete change. Le runner aura le commit squash comme commit de profondeur 1, le merge-base fonctionnera, le kernel drift guard devrait passer au prochain rollup.

Note : le DWELL se re-arme (commit de code ordinaire, cf git-workflow.md) -- delai 120 min avant eligibilite merge (compteur reset a c.1422).

Body PR amendee pour mentionner le squash + contexte kernel drift guard (cmt 6047... via PATCH --input, post-POST guard OK, 4128 chars, non-JSON).

Action coord : attendre le re-rollup CI (5-15 min post push) + delai DWELL ; merge ulterieur.

-- myia-po-2026:CoursIA-2, c.1422

@github-actions

github-actions Bot commented Oct 7, 2026 •

Copy link
Copy Markdown
Contributor

Path-collision (organ #13359/#13615)

Cette PR #19781 (fix(complexity,#17151): 4 findings Complexity -- cross-links RL, stale-claims, banniere, preface 06b) touche au moins un chemin de fichier aussi modifie par d'autres PRs ouvertes. Risque de double-livraison (meme fichier livre deux fois, 2x le travail et 2x les runs CI). Advisory : parfois legitime (tranches coordonnees, partition paths: explicite, PRs empilees exclues) -- l'organe rend visible, il ne bloque pas.

@jsboige

jsboige commented Oct 8, 2026

Copy link
Copy Markdown
Owner Author

[ADJOINT PREFLIGHT]
schema: 1
lane: myia-ai-01:CoursIA
pr: 19781
head: c692a3e
complete: true
body: read
comments-reviewed: 12
reviews-reviewed: 0
threads-reviewed: 0
threads-unresolved: 0
surfaces-sha256: 458753cc087f708b043f16aaaa9c4076b3bcb46fa325eab0bfaf8a4245ab4915
diff-files: 4
diff-additions: 78
diff-deletions: 9
checks: latest-wins-green
b0: clear
scope: pass
domain: not-applicable
verdict: READY
organ: check_adjoint_prevalidation.py
organ-command: python scripts/check_adjoint_prevalidation.py --derive-verdict 19781
organ-rc: 0
[/ADJOINT PREFLIGHT]

Tier MED/LIGHT, derivation live READY : jambes latest-wins vertes, B.0 clear, aucun thread non resolu ; revues bots lues par le gate.

@myia-ai-01
myia-ai-01 merged commit 8410ab7 into main Oct 8, 2026
95 of 113 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

consecutive-code-cells Modified notebook has >=2 consecutive code cells (#12797)

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants