docs(iit,#13410): tranche densite ICT-38 + ICT-8 (deficits minimaux residuels) by jsboige · Pull Request #16434 · jsboige/CoursIA

jsboige · 2026-09-16T17:25:52Z

Grain: DEEP/notebook-python -- lane myia-po-2026:CoursIA -- prev: DEEP/notebook-python #16357

Summary

Tranche densite #13410 de cloture des plus petits deficits IIT/ICT : 2 cellules markdown de lecture chiffree (1 + 1). Chaque nombre cite est lu dans les sorties commitees des notebooks eux-memes. Markdown-only : 22 insertions, 0 deletion, aucune cellule code touchee, aucun output edite.

ICT-38-SLens-SelfLocation — 1189 → 1248 : lecture de l'intervention causale appariee (patch experiment) — exactitude de base de l'hote 1.0000, donneur apparie 0.0742 vs cible aleatoire 0.0625 (proche du hasard), sham 0.1953 (exactitude residuelle ~2,6x celle du donneur apparie, qui perturbe donc le plus) : la localisation de soi ne se reduit pas a la position du bloc reinjecte.

ICT-8-AttractorLandscapesEWS — 1175 → 1245 : lecture de la rampe reelle — bascule a t = 184 446, c_tip = 2.622 au franchissement du pli c_fold ≈ 2.604 ; avant-bascule, variance Kendall tau +0.347 (p = 1.9e-237) et AR1 tau +0.303 (p = 5.8e-182) : les precurseurs EWS montent avant la transition, contrairement au regime plateau (AR1 brute 0.9987 saturee, 0.9386 apres amincissement x50).

Validation

pedagogy_density.py : « All judged notebooks meet the density floor » (0 sous-plancher, post-commit).
detect_markdown_rendering.py --check par notebook : OK 2/2 (rc=0).
Commit 2973a3ae9 : 2 notebooks, 22 insertions, 0 deletion. Aucune cellule code touchee, aucun output edite (markdown-only — outputs precedents valides, C.2/C.3).

G-VAR-3 (variation adjacency)

{"guard_pass": true, "blocking": false, "adjacent": true, "genre": "notebook-python", "prev_genre": "notebook-python", "lane": "myia-po-2026:CoursIA", "prev_source": "merged-sequence", "declared_prev_genre": "notebook-python", "prev_pr": 16357, "reason": "adjacence notebook-python->notebook-python hors liste LIGHT : §2 l'autorise si chaque grain est une substance genument distincte -- jugement non mecanisable, signal advisory uniquement.", "sequence_as_of": "2026-09-16T17:01:24Z"} — advisory uniquement (substance distincte : 2 nouveaux notebooks IIT, md-only, deficits minimaux restants de la serie).

See #13410 (contribution partielle au rollout densite).

🤖 Generated with Claude Code

Markdown-only (22 insertions, 0 deletion) : lecture d'intervention causale appariee sur 38 (0.0742/0.0625/0.1953, base 1.0000) et lecture de la rampe EWS sur 8 (t=184446, c_tip 2.622 vs pli 2.604, tau variance +0.347 p=1.9e-237, tau AR1 +0.303 p=5.8e-182). Densite 1189 -> 1248 et 1175 -> 1245, plancher 1200 (#13410). Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

github-actions · 2026-09-16T17:27:41Z

⚠️ 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 · 2026-09-16T17:28:01Z

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

github-actions · 2026-09-16T17:28:57Z

Golden-Set Execution (H.7 P3)

✅ 8/8 notebooks passed (certified reproducible)

Notebook	Status	Time
2.1-Workflow-ML.ipynb	✅ SUCCESS	3.0s
2.2-Descente-de-gradient.ipynb	✅ SUCCESS	3.7s
2.3-Regression-lineaire-logistique.ipynb	✅ SUCCESS	4.1s
2.4-Arbres-Forets-Ensembles.ipynb	✅ SUCCESS	3.8s
Search-01-StateSpace.ipynb	✅ SUCCESS	3.1s
SL-1-LogicalLearning.ipynb	✅ SUCCESS	2.2s
rl_4_multi_armed_bandits.ipynb	✅ SUCCESS	17.7s
GameTheory-04c-NashExistence-Python.ipynb	✅ SUCCESS	3.1s

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

github-actions · 2026-09-16T17:29:36Z

Notebook PR Validation: PASS

Notebooks checked: 2
Code cells validated: 27
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)

myia-ai-01

Exact-head review `2973a3ae9850b4f8a1d79d5fb82a2dda71bbf099`

Le périmètre markdown-only et tous les nombres sont bien ancrés dans les sorties committées, mais la nouvelle lecture ICT-38 inverse le sens du résultat.

L’exactitude de base vaut 1.0000. Après intervention, le donneur apparié vaut 0.0742, la cible aléatoire 0.0625, et le sham 0.1953. Une exactitude plus basse signifie ici une perturbation plus forte. Le texte conclut pourtant : « le sham perturbe nettement davantage », alors que 0.1953 > 0.0742 montre au contraire que le donneur apparié dégrade davantage l’exactitude que le sham. Le ratio 0.1953 / 0.0742 ≈ 2,6 est arithmétiquement correct mais ne justifie pas la direction causale écrite.

Merci de corriger la comparaison et la conclusion sans toucher au code ni aux outputs, puis de pousser un nouveau head. Toute nouvelle tête réinitialise DWELL ; re-review exacte-head requise. Aucun merge sur cette tête.

Review exact-head #16434 : exactitude basse = perturbation forte, donc le donneur apparie (0.0742) perturbe davantage que le sham (0.1953). Markdown-only 2+/2-, code/outputs intacts. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

…-ict38-8

github-actions · 2026-09-16T19:00:09Z

Path-collision (organ #13359/#13615)

Cette PR #16434 (docs(iit,#13410): tranche densite ICT-38 + ICT-8 (deficits minimaux residuels)) 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.

faible -- docs(iit,#16224): ICT-38/ICT-37 -- cellule « Voisins et positionnement » (antecedents nommes, numpy-only justifie, renvoi ICT-17) #16278 partage : MyIA.AI.Notebooks/IIT/ICT-Series/ICT-38-SLens-SelfLocation.ipynb
fort -- docs(iit,#13410): densite ICT-8 - lectures chiffrees (1248 -> 1493) #16455 partage : MyIA.AI.Notebooks/IIT/ICT-Series/ICT-8-AttractorLandscapesEWS.ipynb (issues communes : densite pedagogique : 430 notebooks sous le plancher 1200 — surface majoritairement non suivie #13410)

myia-ai-01

Approving exact head ac219990d59ecf6017ffb9f6474691293aa25616 after re-reading the complete PR surface and independently verifying the repair, notebook invariants, collisions, current-main integration, and current-head checks.

The prior direction error is corrected: the prose now states that the paired transfer is the stronger perturbation (0.0742 accuracy versus 0.1953 for the sham), while preserving the supported approximately 2.6 ratio.

The head merges canonical main 5a1989a92e2185678763a06a76032b7384cd70e5 as its second parent without touching either target notebook. The effective delta remains two inserted markdown cells only; code, outputs, and execution counts are unchanged. The current-head gate settled 79 content checks green, no open PR collides on either notebook, and closingIssuesReferences is empty so #13410 remains open.

This approval lifts the prior content concern only. Merge still requires the reset 120-minute DWELL (matures at 2026-09-16T20:46:59Z) and PR gate to pass. Do not push again merely to refresh the gate.

jsboige · 2026-09-17T13:26:56Z

[ADJOINT PREFLIGHT]
schema: 1
lane: myia-po-2025:CoursIA-2
pr: 16434
head: ac21999
complete: true
body: read
comments-reviewed: 5
reviews-reviewed: 2
threads-reviewed: 0
threads-unresolved: 0
surfaces-sha256: 3b178e360f46920378046a836cd0a5ca3ca82584fb93640f95fd11cb1f83d1a7
diff-files: 2
diff-additions: 22
diff-deletions: 0
checks: latest-wins-green
b0: clear
scope: pass
domain: pass
verdict: READY
[/ADJOINT PREFLIGHT]

jsboige mentioned this pull request Sep 16, 2026

densite pedagogique : 430 notebooks sous le plancher 1200 — surface majoritairement non suivie #13410

Open

myia-ai-01 requested changes Sep 16, 2026

View reviewed changes

jsboige and others added 2 commits September 16, 2026 20:18

Merge remote-tracking branch 'origin/main' into feature/13410-density…

ac21999

…-ict38-8

github-actions Bot mentioned this pull request Sep 16, 2026

docs(iit,#16224): ICT-38/ICT-37 -- cellule « Voisins et positionnement » (antecedents nommes, numpy-only justifie, renvoi ICT-17) #16278

Merged

myia-ai-01 approved these changes Sep 16, 2026

View reviewed changes

jsboige mentioned this pull request Sep 16, 2026

docs(iit,#13410): densite ICT-13 (1029->1243) + ICT-15f (994->1347) -- 8 lectures chiffrees #16444

Merged

github-actions Bot mentioned this pull request Sep 16, 2026

docs(iit,#13410): densite ICT-8 - lectures chiffrees (1248 -> 1493) #16455

Merged

github-actions Bot added the pr-overlap Advisory: another open PR touches the same files (organ #13615) label Sep 16, 2026

myia-ai-01 merged commit fbcb831 into main Sep 17, 2026
81 of 82 checks passed

This was referenced Sep 17, 2026

fix(infra,#15574): baseline topologique myia-po-2024 — organe cross-platform #16313

Merged

fix(density,#17040): redressement paquet P09 -- series IIT #17060

Merged

Navigation Menu

Repository navigation

docs(iit,#13410): tranche densite ICT-38 + ICT-8 (deficits minimaux residuels) - #16434

docs(iit,#13410): tranche densite ICT-38 + ICT-8 (deficits minimaux residuels)#16434
myia-ai-01 merged 3 commits into
mainfrom
feature/13410-density-ict38-8

jsboige commented Sep 16, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Sep 16, 2026

Uh oh!

github-actions Bot commented Sep 16, 2026

Uh oh!

github-actions Bot commented Sep 16, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Sep 16, 2026

Uh oh!

myia-ai-01 left a comment

Uh oh!

github-actions Bot commented Sep 16, 2026 •

edited

Loading

Uh oh!

myia-ai-01 left a comment

Uh oh!

jsboige commented Sep 17, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

jsboige commented Sep 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Validation

G-VAR-3 (variation adjacency)

Uh oh!

github-actions Bot commented Sep 16, 2026

Uh oh!

github-actions Bot commented Sep 16, 2026

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

Uh oh!

github-actions Bot commented Sep 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Golden-Set Execution (H.7 P3)

Uh oh!

github-actions Bot commented Sep 16, 2026

Notebook PR Validation: PASS

Uh oh!

myia-ai-01 left a comment

Choose a reason for hiding this comment

Exact-head review 2973a3ae9850b4f8a1d79d5fb82a2dda71bbf099

Uh oh!

github-actions Bot commented Sep 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Path-collision (organ #13359/#13615)

Uh oh!

myia-ai-01 left a comment

Choose a reason for hiding this comment

Uh oh!

jsboige commented Sep 17, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

jsboige commented Sep 16, 2026 •

edited

Loading

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

github-actions Bot commented Sep 16, 2026 •

edited

Loading

Exact-head review `2973a3ae9850b4f8a1d79d5fb82a2dda71bbf099`

github-actions Bot commented Sep 16, 2026 •

edited

Loading