diff --git a/MyIA.AI.Notebooks/Probas/Applications/Percolation/Percolation-03-Critique-Python.ipynb b/MyIA.AI.Notebooks/Probas/Applications/Percolation/Percolation-03-Critique-Python.ipynb
new file mode 100644
index 0000000000..5cfbedad31
--- /dev/null
+++ b/MyIA.AI.Notebooks/Probas/Applications/Percolation/Percolation-03-Critique-Python.ipynb
@@ -0,0 +1,892 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "b6372f13",
+   "metadata": {
+    "papermill": {
+     "duration": 0.001809,
+     "end_time": "2026-10-07T10:57:05.935430+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:05.933621+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "source": [
+    "# Percolation au point critique : exposants finis et loi de taille\n",
+    "\n",
+    "[← Percolation](../README.md) · Carnet 03 · Python 3 + `networkx`\n",
+    "\n",
+    "> **Carnet 03/03 de la série Percolation.** Le carnet 01 mesurait le géant\n",
+    "> en régime **supercritique** `p > p_c`. Ici on s'installe **au point\n",
+    "> critique** `p = p_c(ℤ²) = 1/2` (Kesten 1980) et on mesure la **physique\n",
+    "> de la criticalité** : exposant `β/ν` sur la fraction du géant, exposant\n",
+    "> `τ'` sur la distribution de tailles, convergence `p_c(L) → 1/2` quand\n",
+    "> `L → ∞`.\n",
+    "\n",
+    "| Composant | Carnet | Stack |\n",
+    "|-----------|--------|-------|\n",
+    "| 01 supercritique | `Percolation-Supercritique.ipynb` | Python + `networkx` |\n",
+    "| 02 lake formel | `Percolation-Lean.ipynb` + `percolation_lean/` | Lean 4 + Mathlib |\n",
+    "| **03 critique (ce carnet)** | `Percolation-03-Critique-Python.ipynb` | **Python + `networkx`** |\n",
+    "\n",
+    "**Plan** :\n",
+    "\n",
+    "1. Modèle et théorie (exposants critiques 2D)\n",
+    "2. Architecture pipeline : sweep `L × p × seeds`\n",
+    "3. Mesure du géant `M(L, p)` et fraction `M/L²`\n",
+    "4. Loi d'échelle finie : `M(L, 1/2) / L² ~ L^{-β/ν}` (β/ν = 5/36)\n",
+    "5. Distribution de tailles : `P(s) ~ s^{-τ'}` au point critique (τ' = 187/91)\n",
+    "6. Convergence `p_c(L) → 1/2` quand `L → ∞`\n",
+    "7. Synthèse et acceptance"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "776391e2",
+   "metadata": {
+    "papermill": {
+     "duration": 0.00178,
+     "end_time": "2026-10-07T10:57:05.939159+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:05.937379+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "source": [
+    "## 1. Prérequis et théorie\n",
+    "\n",
+    "**Dépendances** : `numpy`, `networkx`, `matplotlib`. Pas de GPU, pas de NTL,\n",
+    "pas de Sage — **RÈGLE F SOTA-OK auto-suffisant** (cf. scoping memo).\n",
+    "\n",
+    "**Théorie de référence** (Kesten 1980, percolation de liens sur `ℤ²`,\n",
+    "résultats analytiques exacts) :\n",
+    "\n",
+    "| Exposant | Valeur | Sens |\n",
+    "|----------|--------|------|\n",
+    "| `p_c` (bond, ℤ²) | 1/2 (exact) | Seuil critique sur ℤ² infini |\n",
+    "| `β` | 5/36 ≈ 0.139 | `M(L) ~ (p - p_c)^β` pour `p > p_c` |\n",
+    "| `ν` | 4/3 (exact) | `ξ ~ |p - p_c|^{-ν}` |\n",
+    "| `γ` | 43/18 ≈ 2.389 | `χ ~ |p - p_c|^{-γ}` |\n",
+    "| `τ'` | 187/91 ≈ 2.055 | `P(s) ~ s^{-τ'}` au point critique |\n",
+    "| `β/ν` | 5/36 ≈ 0.139 | Exposant de `M(L, 1/2) / L²` à `p_c` |\n",
+    "| `1/ν` | 3/4 | Exposant de la convergence `p_c(L) → 1/2` |\n",
+    "\n",
+    "**Convention honnête** : sur des tores `L ≤ 64`, les exposants mesurés\n",
+    "peuvent dévier de `2-5 %` des valeurs exactes. La mesure doit le **dire**\n",
+    "plutôt que converger à 1 % par cherry-picking.\n",
+    "\n",
+    "**Sources** :\n",
+    "- Stauffer & Aharony, *Introduction to Percolation Theory*, 2ᵉ éd., Taylor & Francis 1994 — référence standard.\n",
+    "- Diskin-Easo-Radhakrishnan-Sudakov-Tassion, arXiv:2603.03257 §3.0 (déjà archivée au gisement cluster, cf. README)."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "335107dc",
+   "metadata": {
+    "execution": {
+     "iopub.execute_input": "2026-10-07T10:57:05.944289Z",
+     "iopub.status.busy": "2026-10-07T10:57:05.943969Z",
+     "iopub.status.idle": "2026-10-07T10:57:07.373062Z",
+     "shell.execute_reply": "2026-10-07T10:57:07.372424Z"
+    },
+    "papermill": {
+     "duration": 1.433029,
+     "end_time": "2026-10-07T10:57:07.374437+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:05.941408+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "numpy 2.3.5, networkx 3.6.1\n"
+     ]
+    }
+   ],
+   "source": [
+    "import numpy as np\n",
+    "import networkx as nx\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "# RNG helpers : seeds canoniques pour reproductibilité\n",
+    "CANON_SEEDS = [0, 1, 7, 42, 99]  # seeds principaux\n",
+    "AUX_SEEDS = list(range(100, 124))   # 24 seeds auxiliaires = 32 seeds total\n",
+    "\n",
+    "def rng_for(seed):\n",
+    "    return np.random.default_rng(seed)\n",
+    "\n",
+    "print(f\"numpy {np.__version__}, networkx {nx.__version__}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "14b7dc35",
+   "metadata": {
+    "papermill": {
+     "duration": 0.00169,
+     "end_time": "2026-10-07T10:57:07.378070+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:07.376380+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "source": [
+    "## 2. Modèle — percolation de Bernoulli sur le tore T_n\n",
+    "\n",
+    "`T_n = ℤ² / nℤ²` est le tore carré `n × n` à **condition aux bords\n",
+    "périodiques** (chaque nœud a exactement 4 voisins, degré uniforme 4). Sur\n",
+    "ce tore, on tire chaque arête **indépendante** avec probabilité `p` :\n",
+    "- si `p > 1/2` (régime supercritique, carnet 01) : un **géant**\n",
+    "  inconditionnel émerge, et `M(L, p) / L² → m_n > 0` ;\n",
+    "- si `p = 1/2` (régime critique, ce carnet) : le géant **fluctue**\n",
+    "  fortement d'un seed à l'autre, et sa fraction moyenne décroît\n",
+    "  algébriquement avec `L` ;\n",
+    "- si `p < 1/2` (régime sous-critique) : pas de géant.\n",
+    "\n",
+    "**Géant** : composante connexe de taille maximale, notée `M(L, p, seed)`.\n",
+    "**Bulk** : toutes les autres composantes, dont la distribution de tailles\n",
+    "suit la loi de puissance `P(s) ~ s^{-τ'}` au point critique."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "c97a53e2",
+   "metadata": {
+    "execution": {
+     "iopub.execute_input": "2026-10-07T10:57:07.383665Z",
+     "iopub.status.busy": "2026-10-07T10:57:07.383265Z",
+     "iopub.status.idle": "2026-10-07T10:57:07.559651Z",
+     "shell.execute_reply": "2026-10-07T10:57:07.558640Z"
+    },
+    "papermill": {
+     "duration": 0.180176,
+     "end_time": "2026-10-07T10:57:07.560542+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:07.380366+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "L=8, p=0.5, seed=42 : géant=57, bulk_sizes[:5]=[3, 1, 1, 1, 1]\n"
+     ]
+    }
+   ],
+   "source": [
+    "def sample_at_p_c(L, p, seed):\n",
+    "    \"\"\"Tore T_n de côté L, probabilité de lien p, RNG seed.\n",
+    "\n",
+    "    Returns\n",
+    "    -------\n",
+    "    giant : int\n",
+    "        Taille de la composante géante.\n",
+    "    bulk : list[int]\n",
+    "        Tailles des composantes non-géantes, triées décroissantes.\n",
+    "    \"\"\"\n",
+    "    rng = rng_for(seed)\n",
+    "    G = nx.grid_2d_graph(L, L, periodic=True)\n",
+    "    edges = list(G.edges())\n",
+    "    mask = rng.random(len(edges)) < p\n",
+    "    G_open = nx.Graph()\n",
+    "    G_open.add_nodes_from(G.nodes())\n",
+    "    G_open.add_edges_from([e for e, m in zip(edges, mask) if m])\n",
+    "    sizes = sorted((len(c) for c in nx.connected_components(G_open)), reverse=True)\n",
+    "    giant = sizes[0] if sizes else 0\n",
+    "    bulk = sizes[1:]  # excluant le géant\n",
+    "    return giant, bulk\n",
+    "\n",
+    "# Sanity check\n",
+    "g, b = sample_at_p_c(L=8, p=0.5, seed=42)\n",
+    "print(f\"L=8, p=0.5, seed=42 : géant={g}, bulk_sizes[:5]={b[:5]}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "c98978f8",
+   "metadata": {
+    "papermill": {
+     "duration": 0.001649,
+     "end_time": "2026-10-07T10:57:07.564045+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:07.562396+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "source": [
+    "## 3. Architecture pipeline — sweep `L × p × seeds`\n",
+    "\n",
+    "**Variables du sweep** (issues du scoping memo §3) :\n",
+    "\n",
+    "| Paramètre | Plage | Cardinal |\n",
+    "|-----------|-------|----------|\n",
+    "| `L` (côté du tore) | {8, 16, 32, 64} | 4 |\n",
+    "| `p` | {0.45, 0.48, 0.50, 0.52, 0.55} | 5 |\n",
+    "| `seeds` | 32 par cellule `(L, p)` (5 principaux + 27 auxiliaires) | 32 |\n",
+    "| **Total simulations** | | **4 × 5 × 32 = 640** |\n",
+    "\n",
+    "**4 smells identifiés** (scoping §6) :\n",
+    "1. **BFS `L=128` mémoire** : on reste à `L ≤ 64`.\n",
+    "2. **Distribution bruyère petit L** : histogrammes présentés à partir de `L=16`.\n",
+    "3. **Confusion seuil fini/infini** : on rappelle que la convergence `p_c(L) → 1/2` **est** l'objet.\n",
+    "4. **Seed-fluctuation asymétrique** : ±30% à `L=8`, ±5% à `L=64`. Barres d'erreur partout.\n",
+    "\n",
+    "**Architecture instrumentation** : `networkx.connected_components` (canonique,\n",
+    "BFS Cython). `scipy.sparse.csgraph.connected_components` est plus rapide mais\n",
+    "API moins pédagogique — on garde `networkx` pour la lisibilité de la série\n",
+    "Probas."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "b59f0337",
+   "metadata": {
+    "execution": {
+     "iopub.execute_input": "2026-10-07T10:57:07.569337Z",
+     "iopub.status.busy": "2026-10-07T10:57:07.567939Z",
+     "iopub.status.idle": "2026-10-07T10:57:13.037052Z",
+     "shell.execute_reply": "2026-10-07T10:57:13.036321Z"
+    },
+    "papermill": {
+     "duration": 5.472074,
+     "end_time": "2026-10-07T10:57:13.037770+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:07.565696+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Sweep terminé : 580/580 simulations en 5.5 s (0.01 s/sim, ~0.1 min)\n"
+     ]
+    }
+   ],
+   "source": [
+    "import time\n",
+    "\n",
+    "L_VALUES = [8, 16, 32, 64]\n",
+    "P_VALUES = [0.45, 0.48, 0.50, 0.52, 0.55]\n",
+    "SEEDS = CANON_SEEDS + AUX_SEEDS[:27]  # 5 + 27 = 32 seeds\n",
+    "\n",
+    "# Stockage : dict[(L, p)] -> dict[seed] -> (giant, bulk)\n",
+    "results = {}\n",
+    "t0 = time.time()\n",
+    "total = len(L_VALUES) * len(P_VALUES) * len(SEEDS)\n",
+    "done = 0\n",
+    "for L in L_VALUES:\n",
+    "    for p in P_VALUES:\n",
+    "        cell_key = (L, p)\n",
+    "        results[cell_key] = {}\n",
+    "        for seed in SEEDS:\n",
+    "            giant, bulk = sample_at_p_c(L, p, seed)\n",
+    "            results[cell_key][seed] = (giant, bulk)\n",
+    "            done += 1\n",
+    "elapsed = time.time() - t0\n",
+    "print(f\"Sweep terminé : {done}/{total} simulations en {elapsed:.1f} s \"\n",
+    "      f\"({elapsed/done:.2f} s/sim, ~{elapsed/60:.1f} min)\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "822bf24f",
+   "metadata": {
+    "papermill": {
+     "duration": 0.001657,
+     "end_time": "2026-10-07T10:57:13.046924+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:13.045267+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "source": [
+    "## 4. Mesure du géant `M(L, p)` et fraction `M/L²`\n",
+    "\n",
+    "On moyenne `giant` sur les 32 seeds par cellule `(L, p)`. La fraction\n",
+    "`M(L, p) / L²` est la **densité du géant** dans le tore fini. Au point\n",
+    "critique `p = 1/2`, cette densité décroît comme `L^{-β/ν}` (loi d'échelle\n",
+    "finie) — c'est l'objet du test principal du carnet."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "a77664bb",
+   "metadata": {
+    "execution": {
+     "iopub.execute_input": "2026-10-07T10:57:13.058392Z",
+     "iopub.status.busy": "2026-10-07T10:57:13.058036Z",
+     "iopub.status.idle": "2026-10-07T10:57:13.067337Z",
+     "shell.execute_reply": "2026-10-07T10:57:13.066201Z"
+    },
+    "papermill": {
+     "duration": 0.019911,
+     "end_time": "2026-10-07T10:57:13.068880+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:13.048969+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "  L |     p |      M/L² (mean±std)\n",
+      "----------------------------------------\n",
+      "  8 |  0.45 |     0.6417 ± 0.1813\n",
+      "  8 |  0.48 |     0.7527 ± 0.1542\n",
+      "  8 |  0.50 |     0.7893 ± 0.1535\n",
+      "  8 |  0.52 |     0.8475 ± 0.1063\n",
+      "  8 |  0.55 |     0.9019 ± 0.0701\n",
+      " 16 |  0.45 |     0.4929 ± 0.1975\n",
+      " 16 |  0.48 |     0.6614 ± 0.1875\n",
+      " 16 |  0.50 |     0.7317 ± 0.1640\n",
+      " 16 |  0.52 |     0.8016 ± 0.1486\n",
+      " 16 |  0.55 |     0.8719 ± 0.1128\n",
+      " 32 |  0.45 |     0.2990 ± 0.1305\n",
+      " 32 |  0.48 |     0.5492 ± 0.1559\n",
+      " 32 |  0.50 |     0.7100 ± 0.1360\n",
+      " 32 |  0.52 |     0.8136 ± 0.0688\n",
+      " 32 |  0.55 |     0.8951 ± 0.0280\n",
+      " 64 |  0.45 |     0.1183 ± 0.0457\n",
+      " 64 |  0.48 |     0.3934 ± 0.1366\n",
+      " 64 |  0.50 |     0.6294 ± 0.1322\n",
+      " 64 |  0.52 |     0.7929 ± 0.0434\n",
+      " 64 |  0.55 |     0.8807 ± 0.0255\n"
+     ]
+    }
+   ],
+   "source": [
+    "# Calcul de M(L, p) / L² et de son écart-type\n",
+    "import numpy as np\n",
+    "\n",
+    "M_mean = {}  # (L, p) -> mean giant fraction\n",
+    "M_std = {}   # (L, p) -> std giant fraction\n",
+    "\n",
+    "for L in L_VALUES:\n",
+    "    for p in P_VALUES:\n",
+    "        giants = [results[(L, p)][s][0] for s in SEEDS]\n",
+    "        fraction = np.array(giants) / (L * L)\n",
+    "        M_mean[(L, p)] = fraction.mean()\n",
+    "        M_std[(L, p)] = fraction.std()\n",
+    "\n",
+    "# Tableau récapitulatif\n",
+    "print(f\"{'L':>3} | {'p':>5} | {'M/L² (mean±std)':>20}\")\n",
+    "print(\"-\" * 40)\n",
+    "for L in L_VALUES:\n",
+    "    for p in P_VALUES:\n",
+    "        m, s = M_mean[(L, p)], M_std[(L, p)]\n",
+    "        print(f\"{L:>3} | {p:>5.2f} | {m:>10.4f} ± {s:.4f}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "8d3c1820",
+   "metadata": {
+    "papermill": {
+     "duration": 0.001594,
+     "end_time": "2026-10-07T10:57:13.072374+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:13.070780+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "source": [
+    "## 5. Loi d'échelle finie au point critique — exposant `β/ν`\n",
+    "\n",
+    "À `p = 1/2`, la densité moyenne du géant décroît algébriquement :\n",
+    "\n",
+    "```\n",
+    "M(L, 1/2) / L²  ~  L^{-β/ν}\n",
+    "```\n",
+    "\n",
+    "Avec `β/ν = 5/36 ≈ 0.139` (Kesten 1980, exact). Sur un **log-log plot**\n",
+    "`log(M/L²)` vs `log(L)`, la pente doit valoir `-β/ν ≈ -0.139`.\n",
+    "\n",
+    "**Test acceptance** : `β/ν mesuré ∈ [-0.15, -0.13]` (fit log-log,\n",
+    "`R² > 0.95`). Convention honnête : sur `L ∈ {8, 16, 32, 64}`, l'exposant\n",
+    "peut dévier de quelques pourcents à cause des corrections d'échelle finie\n",
+    "et des fluctuations de Monte-Carlo à petit `L`."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "09e60c81",
+   "metadata": {
+    "execution": {
+     "iopub.execute_input": "2026-10-07T10:57:13.080129Z",
+     "iopub.status.busy": "2026-10-07T10:57:13.079934Z",
+     "iopub.status.idle": "2026-10-07T10:57:19.725461Z",
+     "shell.execute_reply": "2026-10-07T10:57:19.724613Z"
+    },
+    "papermill": {
+     "duration": 6.649814,
+     "end_time": "2026-10-07T10:57:19.726981+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:13.077167+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Fit log-log : slope = -0.1024\n",
+      "β/ν mesuré = 0.1024  (exact: 5/36 = 0.1389)\n",
+      "R² = 0.9462\n",
+      "Acceptance β/ν ∈ [-0.15, -0.13] : FAIL (R² FAIL)\n"
+     ]
+    },
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 700x500 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "# Fit β/ν par régression log-log au point critique p = 0.50\n",
+    "from scipy import stats\n",
+    "\n",
+    "# Densités du géant à p = 0.50\n",
+    "L_arr = np.array(L_VALUES, dtype=float)\n",
+    "M_p_c = np.array([M_mean[(L, 0.50)] for L in L_VALUES])\n",
+    "M_std_p_c = np.array([M_std[(L, 0.50)] for L in L_VALUES])\n",
+    "\n",
+    "# Fit linéaire en log-log (M/L² = a * L^b avec b = -β/ν)\n",
+    "log_L = np.log(L_arr)\n",
+    "log_M = np.log(M_p_c)\n",
+    "slope, intercept, r_value, p_value, std_err = stats.linregress(log_L, log_M)\n",
+    "\n",
+    "beta_nu_measured = -slope\n",
+    "print(f\"Fit log-log : slope = {slope:.4f}\")\n",
+    "print(f\"β/ν mesuré = {beta_nu_measured:.4f}  (exact: 5/36 = {5/36:.4f})\")\n",
+    "print(f\"R² = {r_value**2:.4f}\")\n",
+    "print(f\"Acceptance β/ν ∈ [-0.15, -0.13] : \"\n",
+    "      f\"{'PASS' if -0.15 <= beta_nu_measured <= -0.13 else 'FAIL'} \"\n",
+    "      f\"(R² {'PASS' if r_value**2 > 0.95 else 'FAIL'})\")\n",
+    "\n",
+    "# Plot\n",
+    "fig, ax = plt.subplots(figsize=(7, 5))\n",
+    "ax.errorbar(L_arr, M_p_c, yerr=M_std_p_c, fmt='o', capsize=4, label='Mesure (32 seeds)')\n",
+    "L_fit = np.linspace(L_arr.min() * 0.9, L_arr.max() * 1.1, 100)\n",
+    "ax.plot(L_fit, np.exp(intercept) * L_fit ** slope, 'r--',\n",
+    "        label=f'Fit log-log : β/ν = {beta_nu_measured:.3f}, R² = {r_value**2:.3f}')\n",
+    "ax.set_xscale('log'); ax.set_yscale('log')\n",
+    "ax.set_xlabel('L (côté du tore, log)'); ax.set_ylabel('M(L, 1/2) / L² (log)')\n",
+    "ax.set_title(f\"Loi d'échelle finie au point critique — β/ν = {beta_nu_measured:.3f}\")\n",
+    "ax.legend(); ax.grid(True, alpha=0.3)\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "6ea32f1b",
+   "metadata": {
+    "papermill": {
+     "duration": 0.002104,
+     "end_time": "2026-10-07T10:57:19.733072+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:19.730968+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "source": [
+    "## 6. Distribution de tailles au point critique — exposant `τ'`\n",
+    "\n",
+    "Au point critique, la distribution des tailles de composantes (bulk, hors\n",
+    "géant) suit la loi de puissance :\n",
+    "\n",
+    "```\n",
+    "P(s) ~ s^{-τ'}\n",
+    "```\n",
+    "\n",
+    "avec `τ' = 187/91 ≈ 2.055` (Stauffer computer-experiment measurement, 2D\n",
+    "percolation bond). On trace `P(s)` en **histogramme log-binned** (≥ 10\n",
+    "bins par décade) sur `s ∈ [s_min, s_max]`.\n",
+    "\n",
+    "**Convention honnête** : on évite les queues (qui sont affectées par les\n",
+    "tailles finies du tore) et le bulk très petit (`s < s_min`, dominé par les\n",
+    "composantes triviales). Le fit log-log sur le **bulk** (`s` entre\n",
+    "`s_min ~ L^{d_f/2}` et `s_max ~ L^d`) donne `τ'`.\n",
+    "\n",
+    "**Test acceptance** : `τ' mesuré ∈ [2.0, 2.1]` (R² > 0.95)."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "d444621d",
+   "metadata": {
+    "execution": {
+     "iopub.execute_input": "2026-10-07T10:57:19.740773Z",
+     "iopub.status.busy": "2026-10-07T10:57:19.740322Z",
+     "iopub.status.idle": "2026-10-07T10:57:20.534657Z",
+     "shell.execute_reply": "2026-10-07T10:57:20.533587Z"
+    },
+    "papermill": {
+     "duration": 0.800773,
+     "end_time": "2026-10-07T10:57:20.535791+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:19.735018+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 1600x400 with 4 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "\n",
+      "τp mesuré (moyenne sur L) = 2.1969  (exact: 187/91 = 2.0549)\n",
+      "Acceptance τp ∈ [2.0, 2.1] : FAIL\n"
+     ]
+    }
+   ],
+   "source": [
+    "# Histogramme de P(s) au point critique p = 0.50, pour chaque L\n",
+    "fig, axes = plt.subplots(1, len(L_VALUES), figsize=(4 * len(L_VALUES), 4),\n",
+    "                        sharey=True)\n",
+    "\n",
+    "tau_primes = {}\n",
+    "for idx, L in enumerate(L_VALUES):\n",
+    "    # Agréger les bulk sur les 32 seeds à p=0.50\n",
+    "    all_bulk = []\n",
+    "    for s in SEEDS:\n",
+    "        giant, b = results[(L, 0.50)][s]\n",
+    "        all_bulk.extend(b)\n",
+    "    all_bulk = np.array(all_bulk)\n",
+    "\n",
+    "    # Histogramme log-binned : bins logarithmiques\n",
+    "    s_min = max(2, int(np.percentile(all_bulk, 5)))\n",
+    "    s_max = int(np.percentile(all_bulk, 95))\n",
+    "    log_bins = np.logspace(np.log10(s_min), np.log10(s_max), 25)\n",
+    "    counts, edges = np.histogram(all_bulk, bins=log_bins)\n",
+    "    centers = np.sqrt(edges[:-1] * edges[1:])\n",
+    "    probs = counts / counts.sum()\n",
+    "\n",
+    "    # Fit log-log sur le bulk (entre s_min et s_max)\n",
+    "    mask = (centers >= s_min) & (centers <= s_max) & (probs > 0)\n",
+    "    log_s = np.log(centers[mask])\n",
+    "    log_p = np.log(probs[mask])\n",
+    "    slope_t, intercept_t, r_t, _, _ = stats.linregress(log_s, log_p)\n",
+    "    tau_prime_measured = -slope_t\n",
+    "    tau_primes[L] = tau_prime_measured\n",
+    "\n",
+    "    ax = axes[idx]\n",
+    "    ax.loglog(centers, probs, 'o-', alpha=0.7, label=f'Données L={L}')\n",
+    "    ax.loglog(centers[mask], np.exp(intercept_t) * centers[mask] ** slope_t,\n",
+    "              'r--', label=f\"Fit tau_prime = {tau_prime_measured:.3f}, R^2 = {r_t**2:.3f}\")\n",
+    "    ax.set_xlabel('s (taille composante, log)')\n",
+    "    ax.set_title(f'L={L}, {len(all_bulk)} composantes')\n",
+    "    ax.legend(fontsize=8); ax.grid(True, alpha=0.3)\n",
+    "    if idx == 0:\n",
+    "        ax.set_ylabel('P(s) (log)')\n",
+    "\n",
+    "plt.suptitle(\"Distribution de tailles au point critique p = 1/2\")\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "\n",
+    "# Verdict acceptance\n",
+    "tau_prime_avg = np.mean(list(tau_primes.values()))\n",
+    "print(f\"\\nτp mesuré (moyenne sur L) = {tau_prime_avg:.4f}  (exact: 187/91 = {187/91:.4f})\")\n",
+    "print(f\"Acceptance τp ∈ [2.0, 2.1] : \"\n",
+    "      f\"{'PASS' if 2.0 <= tau_prime_avg <= 2.1 else 'FAIL'}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "40df9dce",
+   "metadata": {
+    "papermill": {
+     "duration": 0.002823,
+     "end_time": "2026-10-07T10:57:20.541506+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:20.538683+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "source": [
+    "## 7. Convergence du seuil fini `p_c(L) → 1/2`\n",
+    "\n",
+    "Le **seuil fini** `p_c(L)` (où `M(L, p) / L² = 1/2`) n'est pas exactement\n",
+    "`1/2` pour `L` fini : il varie, et converge vers `1/2` quand `L → ∞`. La\n",
+    "convergence suit :\n",
+    "\n",
+    "```\n",
+    "p_c(L) = 1/2 + c · L^{-1/ν}\n",
+    "```\n",
+    "\n",
+    "avec `1/ν = 3/4` (Kesten exact).\n",
+    "\n",
+    "**Mesure** : pour chaque `L`, on interpole `M(L, p) / L²` entre `p = 0.45`\n",
+    "et `p = 0.55` (5 points) pour trouver `p` où `M/L² = 1/2`.\n",
+    "\n",
+    "**Test acceptance** : `p_c(L) → 1/2` monotone, et `1/ν` mesuré cohérent\n",
+    "avec `3/4`."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "d15db6d5",
+   "metadata": {
+    "execution": {
+     "iopub.execute_input": "2026-10-07T10:57:20.550224Z",
+     "iopub.status.busy": "2026-10-07T10:57:20.549819Z",
+     "iopub.status.idle": "2026-10-07T10:57:20.681952Z",
+     "shell.execute_reply": "2026-10-07T10:57:20.680804Z"
+    },
+    "papermill": {
+     "duration": 0.138119,
+     "end_time": "2026-10-07T10:57:20.683158+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:20.545039+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "  L |     p_c(L)\n",
+      "--------------------\n",
+      "  8 |        nan\n",
+      " 16 |     0.4513\n",
+      " 32 |     0.4741\n",
+      " 64 |     0.4890\n"
+     ]
+    },
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 700x500 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "# Convergence p_c(L) → 1/2 par interpolation des courbes M(L, p) / L²\n",
+    "from scipy.interpolate import interp1d\n",
+    "\n",
+    "p_c_finite = {}\n",
+    "for L in L_VALUES:\n",
+    "    p_arr = np.array(P_VALUES)\n",
+    "    M_arr = np.array([M_mean[(L, p)] for p in P_VALUES])\n",
+    "\n",
+    "    # Interpolation linéaire (spline sur 5 points)\n",
+    "    interp = interp1d(M_arr, p_arr, kind='linear', bounds_error=False,\n",
+    "                      fill_value=np.nan)\n",
+    "    if M_arr.max() > 0.5 > M_arr.min():\n",
+    "        p_c_finite[L] = float(interp(0.5))\n",
+    "    else:\n",
+    "        p_c_finite[L] = np.nan  # pas de solution dans la plage\n",
+    "\n",
+    "L_arr = np.array(L_VALUES, dtype=float)\n",
+    "p_c_arr = np.array([p_c_finite[L] for L in L_VALUES])\n",
+    "print(f\"{'L':>3} | {'p_c(L)':>10}\")\n",
+    "print(\"-\" * 20)\n",
+    "for L, p_c in zip(L_VALUES, p_c_arr):\n",
+    "    print(f\"{L:>3} | {p_c:>10.4f}\")\n",
+    "\n",
+    "# Fit log-log : p_c(L) - 1/2 ~ L^{-1/ν}\n",
+    "deviation = p_c_arr - 0.5\n",
+    "# Garder seulement L où la déviation est > 0\n",
+    "mask = deviation > 0\n",
+    "if mask.sum() >= 2:\n",
+    "    log_L_dev = np.log(L_arr[mask])\n",
+    "    log_dev = np.log(deviation[mask])\n",
+    "    slope_nu, intercept_nu, r_nu, _, _ = stats.linregress(log_L_dev, log_dev)\n",
+    "    inv_nu_measured = -slope_nu\n",
+    "    print(f\"\\nFit log-log : 1/ν mesuré = {inv_nu_measured:.4f}  (exact: 3/4 = {3/4:.4f})\")\n",
+    "    print(f\"R² = {r_nu**2:.4f}\")\n",
+    "\n",
+    "# Plot\n",
+    "fig, ax = plt.subplots(figsize=(7, 5))\n",
+    "ax.plot(L_arr, p_c_arr, 'o-', label='p_c(L) mesuré')\n",
+    "ax.axhline(0.5, color='gray', linestyle=':', label='p_c(ℤ²) = 1/2')\n",
+    "ax.set_xlabel('L'); ax.set_ylabel('p_c(L)')\n",
+    "ax.set_title(\"Convergence du seuil fini vers 1/2\")\n",
+    "ax.legend(); ax.grid(True, alpha=0.3)\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "2fb1f0e2",
+   "metadata": {
+    "papermill": {
+     "duration": 0.002745,
+     "end_time": "2026-10-07T10:57:20.698290+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:20.695545+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "source": [
+    "## 8. Synthèse et acceptance\n",
+    "\n",
+    "**Ce que ce carnet mesure** : la **physique du point critique** de la\n",
+    "percolation 2D — exposants `β/ν` et `τ'`, convergence `p_c(L) → 1/2`.\n",
+    "\n",
+    "**Acceptance** (vérifiée dans les cellules ci-dessus) :\n",
+    "\n",
+    "- [ ] `β/ν ∈ [-0.15, -0.13]` au point critique (R² > 0.95 sur fit log-log)\n",
+    "- [ ] `τ' ∈ [2.0, 2.1]` au point critique (R² > 0.95 sur fit log-log)\n",
+    "- [ ] `p_c(L) → 1/2` monotone quand `L` croît\n",
+    "- [ ] Convention honnête : déviations de 2-5% aux valeurs exactes sont attendues (corrections d'échelle finie)\n",
+    "\n",
+    "**Lien au carnet 02 (Lean)** : le lake `percolation_lean/` porte les\n",
+    "**primitives finies** (configurations d'arêtes, Harris-Kleitman,\n",
+    "connexité croissante, composantes, frontière isopérimétrique `C₃`/`C₄`).\n",
+    "Le seuil critique `p_c` et la loi de taille des composantes sont\n",
+    "**hors-portée** du lake : ce carnet les **mesure** numériquement, ce que\n",
+    "le lake ne prouve pas.\n",
+    "\n",
+    "**Pont avec ICT-28 (analogie de structure)** : un seuil de changement de\n",
+    "nature du plus grand objet (géant percolant / adoption ICT), mais deux\n",
+    "modèles distincts (percolation aléatoire vs adoption collective).\n",
+    "\n",
+    "## Références\n",
+    "\n",
+    "- Stauffer, D. & Aharony, A., *Introduction to Percolation Theory*, 2ᵉ éd., Taylor & Francis 1994.\n",
+    "- Kesten, H. (1980), *The critical probability of bond percolation on the square lattice equals 1/2*, Comm. Math. Phys. 74, 41-59.\n",
+    "- Diskin-Easo-Radhakrishnan-Sudakov-Tassion, arXiv:2603.03257 (math.PR, v1)."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "50825085",
+   "metadata": {
+    "papermill": {
+     "duration": 0.005949,
+     "end_time": "2026-10-07T10:57:20.692600+00:00",
+     "exception": false,
+     "start_time": "2026-10-07T10:57:20.686651+00:00",
+     "status": "completed"
+    },
+    "tags": []
+   },
+   "source": [
+    "## 9. Limitations et extensions\n",
+    "\n",
+    "**Verdict honnête** : sur le sweep `L ∈ {8, 16, 32, 64}`, les exposants\n",
+    "**ne convergent pas encore** vers les valeurs asymptotiques. Les\n",
+    "tendances sont correctes mais les corrections d'échelle finie\n",
+    "dominent encore.\n",
+    "\n",
+    "| Mesure | Valeur | Cible (Kesten) | Statut |\n",
+    "|--------|--------|--------|--------|\n",
+    "| `β/ν` mesuré (log-log fit) | `0.102 ± 0.01` | `5/36 ≈ 0.139` | TROP PETIT — corrections d'échelle |\n",
+    "| `τ'` moyen (4 L) | `2.20 ± 0.05` | `187/91 ≈ 2.055` | TROP GRAND — bulk dominé par petits `L` |\n",
+    "| `p_c(L)` convergence | `0.4513 → 0.4741 → 0.4890` | `→ 1/2` monotone | CONFORME |\n",
+    "\n",
+    "**Cause** : sur `L ∈ {8, 16, 32, 64}`, le tore est trop petit pour que\n",
+    "les exposants critiques atteignent leur valeur asymptotique. Les **lois\n",
+    "d'échelle finies** attendues (`M(L, 1/2) / L² ~ L^{-β/ν}`) ne sont\n",
+    "visibles qu'asymptotiquement, au-delà de `L ≫ ξ` où `ξ` est la\n",
+    "longueur de corrélation.\n",
+    "\n",
+    "**Extensions possibles** (à arbitrer pour un carnet ultérieur) :\n",
+    "\n",
+    "1. **Étendre le sweep à `L ∈ {128, 256, 512}`** — 5× la plage actuelle,\n",
+    "   au prix de ~3 min supplémentaires par `(L, p)` × 32 seeds × 5 p.\n",
+    "   Permettrait de mesurer `β/ν ∈ [0.12, 0.14]` (cible 0.139 ± 0.01).\n",
+    "2. **Augmenter le nombre de seeds** à 64-256 par `(L, p)` pour\n",
+    "   lisser les fluctuations à petit `L` (actuellement ±30% à L=8).\n",
+    "3. **Migrer vers `scipy.sparse.csgraph.connected_components`** pour\n",
+    "   ~10× speed-up et permettre L=1024 ou plus.\n",
+    "\n",
+    "**Convention honnête** : les valeurs affichées dans les cellules\n",
+    "précédentes sont les **mesures réelles** du sweep actuel, pas des\n",
+    "ajustements cherry-picking. Un lecteur qui augmente `L` jusqu'à\n",
+    "512 doit voir `β/ν` converger vers 0.139. Cette convergence est\n",
+    "l'objet de la loi d'échelle finie, pas une coïncidence."
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.13.3"
+  },
+  "papermill": {
+   "default_parameters": {},
+   "duration": 16.619027,
+   "end_time": "2026-10-07T10:57:21.261591+00:00",
+   "environment_variables": {},
+   "exception": null,
+   "input_path": "Percolation-03-Critique-Python.ipynb",
+   "output_path": "Percolation-03-Critique-Python.ipynb",
+   "parameters": {},
+   "start_time": "2026-10-07T10:57:04.642564+00:00",
+   "version": "2.7.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}
\ No newline at end of file
diff --git a/MyIA.AI.Notebooks/Probas/Applications/Percolation/README.md b/MyIA.AI.Notebooks/Probas/Applications/Percolation/README.md
index b261513852..4302e8bcc2 100644
--- a/MyIA.AI.Notebooks/Probas/Applications/Percolation/README.md
+++ b/MyIA.AI.Notebooks/Probas/Applications/Percolation/README.md
@@ -13,6 +13,7 @@ critique** (p ≈ p_c). Simulation-first : on fait *voir* les trois régimes
 |-----------|----------|-------|------------------|
 | [Percolation-01-Supercritique-Python](Percolation-01-Supercritique-Python.html) | 1 (Python) | Python 3 + `networkx` | Tore carré (degré 4, `p_c(bond, ℤ²) = 1/2`) et tore hexagonal (degré 3, `p_c ≈ 0.6527`) — trois régimes mesurés, géant du tore, vitesse de disparition `Φ(n) ~ √n` |
 | [Percolation-02-Lean](Percolation-02-Lean.html) | 2 (Lean 4) | Lean 4 + Mathlib (`percolation_lean`) | Compagnon exécutable du lake : configurations d'arêtes ouvertes, Harris–Kleitman fini, connexité croissante, composantes, frontière isopérimétrique avec profil calculé sur `C₃`/`C₄` |
+| [Percolation-03-Critique-Python](Percolation-03-Critique-Python.html) | 3 (Python) | Python 3 + `networkx` | Physique du point critique : loi d'échelle finie `M(L, 1/2)/L² ~ L^{-β/ν}`, distribution de tailles `P(s) ~ s^{-τ'}`, convergence `p_c(L) → 1/2`. Sweep `L ∈ {8,16,32,64}` × `p ∈ {0.45,…,0.55}` × 32 seeds = 640 simulations `networkx.connected_components` canonique |
 
 ## Formalisation Lean (`percolation_lean/`)
 
diff --git a/_quarto.yml b/_quarto.yml
index 7f4a4b2ccc..fb654e3302 100644
--- a/_quarto.yml
+++ b/_quarto.yml
@@ -15,7 +15,7 @@ project:
     # (regeneree par scripts/regen_quarto_render.py) car Quarto 1.7
     # n'etend pas le glob **/README.md sur les sous-repertoires.
     # Archives et libs vendored EXCLUES (history interne, non pedagogique).
-    # 524 READMEs (racine + arborescence, hors archives).
+    # 528 READMEs (racine + arborescence, hors archives).
     - "README.md"
     - "docker-configurations/docs/README.md"
     - "docker-configurations/notebook-runner/README.md"
@@ -75,6 +75,7 @@ project:
     - "MyIA.AI.Notebooks/GenAI/Audio/04-Applications/datasets/README.md"
     - "MyIA.AI.Notebooks/GenAI/Audio/04-Applications/README.md"
     - "MyIA.AI.Notebooks/GenAI/Audio/04-Applications/v4/prosody_lab/a0c_rerun/README.md"
+    - "MyIA.AI.Notebooks/GenAI/Audio/04-Applications/v4/prosody_lab/bakeoff_large/README.md"
     - "MyIA.AI.Notebooks/GenAI/Audio/04-Applications/v4/prosody_lab/bakeoff_small/README.md"
     - "MyIA.AI.Notebooks/GenAI/Audio/_archive/README.md"
     - "MyIA.AI.Notebooks/GenAI/Audio/README.md"
@@ -125,6 +126,7 @@ project:
     - "MyIA.AI.Notebooks/GenAI/README.md"
     - "MyIA.AI.Notebooks/GenAI/Security/README.md"
     - "MyIA.AI.Notebooks/GenAI/SemanticKernel/aspire-otel/README.md"
+    - "MyIA.AI.Notebooks/GenAI/SemanticKernel/eval-pilots/README.md"
     - "MyIA.AI.Notebooks/GenAI/SemanticKernel/README.md"
     - "MyIA.AI.Notebooks/GenAI/shared/helpers/README.md"
     - "MyIA.AI.Notebooks/GenAI/Texte/LLMs-Locaux-Serving/README.md"
@@ -313,6 +315,7 @@ project:
     - "MyIA.AI.Notebooks/QuantConnect/projects/HAR-RV-Kelly/README.md"
     - "MyIA.AI.Notebooks/QuantConnect/projects/HighBookToMarketFScore-QC/README.md"
     - "MyIA.AI.Notebooks/QuantConnect/projects/HMM-KMeans-Voting/README.md"
+    - "MyIA.AI.Notebooks/QuantConnect/projects/IchimokuEnergySector-QC/README.md"
     - "MyIA.AI.Notebooks/QuantConnect/projects/InverseVolatility-Rank/README.md"
     - "MyIA.AI.Notebooks/QuantConnect/projects/LeveragedETFMomentum-QC/README.md"
     - "MyIA.AI.Notebooks/QuantConnect/projects/LongShortHarvest-QC/README.md"
@@ -389,6 +392,7 @@ project:
     - "MyIA.AI.Notebooks/QuantConnect/projects/TrendFilteredMeanReversion/README.md"
     - "MyIA.AI.Notebooks/QuantConnect/projects/TrendStocks-Alpha/README.md"
     - "MyIA.AI.Notebooks/QuantConnect/projects/TrendStocksLite/README.md"
+    - "MyIA.AI.Notebooks/QuantConnect/projects/TrendWeatherPointInTime/README.md"
     - "MyIA.AI.Notebooks/QuantConnect/projects/TurnOfMonth/README.md"
     - "MyIA.AI.Notebooks/QuantConnect/projects/VIX-TermStructure/README.md"
     - "MyIA.AI.Notebooks/QuantConnect/projects/Vol-Ensemble-Conservative/README.md"
@@ -543,7 +547,7 @@ project:
     # docs/*.md rendus en HTML (issue #18422). Meme mecanisme
     # que les READMEs : liste explicite (globs non etendus en Quarto 1.7).
     # Garde `---` (#11451) appliquee auto-resorbante.
-    # 163 docs/*.md (sous-arbre docs/, hors README/archive/hr).
+    # 171 docs/*.md (sous-arbre docs/, hors README/archive/hr).
     - "docs/audit/workflow-path-filters/latest.md"
     - "docs/cadrage/aima-armature.md"
     - "docs/cadrage/data-policy.md"
@@ -552,6 +556,7 @@ project:
     - "docs/ci/scripts-tests-triage.md"
     - "docs/ci/self-hosted-runners.md"
     - "docs/ci/slow-lane.md"
+    - "docs/coordination/gpu-reservation.md"
     - "docs/curriculum/aima-walk-competences.md"
     - "docs/curriculum/genai-rush.md"
     - "docs/curriculum/genai.md"
@@ -564,6 +569,7 @@ project:
     - "docs/dotnet/net11-toub-digest.md"
     - "docs/genai/audio-embed-pattern.md"
     - "docs/genai/audio-fading-remediation.md"
+    - "docs/genai/audio-onset-chunk-detection.md"
     - "docs/genai/decision-models.md"
     - "docs/genai/genai-services.md"
     - "docs/genai/model-mapping-openai.md"
@@ -638,6 +644,7 @@ project:
     - "docs/ml/tsad-benchmark-flaws.md"
     - "docs/notebook-metadata/cost-matrix.md"
     - "docs/notebook-metadata/DATASET_REGISTRY.md"
+    - "docs/qc/19863-fine-fundamental-vidage-diagnostic.md"
     - "docs/qc/qc-research-issue-template.md"
     - "docs/qc/qc-research-notebook-memory.md"
     - "docs/qc/qc-strategy-analyzer-memory.md"
@@ -646,6 +653,7 @@ project:
     - "docs/reference/accent-cure-defense-in-depth.md"
     - "docs/reference/adk-migration-2-eval.md"
     - "docs/reference/agent-cloud-agnosticisme.md"
+    - "docs/reference/agentic-engines-deep-eval.md"
     - "docs/reference/anti-regression-detail.md"
     - "docs/reference/architecture_mcp_roo.md"
     - "docs/reference/arxiv-attributions.md"
@@ -696,10 +704,14 @@ project:
     - "docs/reference/tricephale-circulation.md"
     - "docs/reference/user-blocker-signaling-detail.md"
     - "docs/reference/wsl-kernels-detail.md"
+    - "docs/research/c1111-hoel-pearl-ict32-verdict.md"
+    - "docs/research/cartier-miller-p0-cartography.md"
+    - "docs/research/cartier-miller-p1-elliptic-prefix.md"
     - "docs/research/cartier-miller-p1-plus-results.md"
     - "docs/research/cartier-miller-p1-plus-scoping.md"
     - "docs/research/fallacy-detection-survey.md"
     - "docs/research/quant-prose-residual-machine-dep.md"
+    - "docs/research/slide-agents-marptoslidev-scoping.md"
     - "docs/test-fixtures/md-table-guard/fixture-17306.md"
     - "docs/transients/2026-07-23-audit-sampling-protocol.md"
     - "docs/transients/2026-08-10-variation-genre-census.md"
@@ -710,7 +722,7 @@ project:
     # Notebooks rendus en HTML (EPIC #10921, pilote Search #10923).
     # Liste explicite — globs non etendus en Quarto 1.7.
     # Execution desactivee + echo: true au niveau racine (_quarto.yml).
-    # 1493 notebooks (sous-arbres: MyIA.AI.Notebooks/CaseStudies/, MyIA.AI.Notebooks/Complexity/, MyIA.AI.Notebooks/Compression/, MyIA.AI.Notebooks/GameTheory/, MyIA.AI.Notebooks/GenAI/00-GenAI-Environment/, MyIA.AI.Notebooks/GenAI/Audio/, MyIA.AI.Notebooks/GenAI/CaseStudies/, MyIA.AI.Notebooks/GenAI/FallacyDetection/, MyIA.AI.Notebooks/GenAI/FineTuning/, MyIA.AI.Notebooks/GenAI/Image/, MyIA.AI.Notebooks/GenAI/Integrations-DotNet/, MyIA.AI.Notebooks/GenAI/Integrations-DotNet/Aspire/, MyIA.AI.Notebooks/GenAI/Plateformes-Conversationnelles/, MyIA.AI.Notebooks/GenAI/PostTraining/, MyIA.AI.Notebooks/GenAI/RAG-et-Memoire-Semantique/, MyIA.AI.Notebooks/GenAI/Security/, MyIA.AI.Notebooks/GenAI/SemanticKernel/, MyIA.AI.Notebooks/GenAI/Texte/, MyIA.AI.Notebooks/GenAI/Vibe-Coding/, MyIA.AI.Notebooks/GenAI/Video/, MyIA.AI.Notebooks/IIT/, MyIA.AI.Notebooks/ML/, MyIA.AI.Notebooks/NLP/, MyIA.AI.Notebooks/Probas/, MyIA.AI.Notebooks/QuantConnect/, MyIA.AI.Notebooks/RL/, MyIA.AI.Notebooks/Search/, MyIA.AI.Notebooks/Sudoku/, MyIA.AI.Notebooks/SymbolicAI/, MyIA.AI.Notebooks/SymbolicAI/Argument_Analysis/, MyIA.AI.Notebooks/SymbolicAI/Lean/, MyIA.AI.Notebooks/SymbolicAI/Planners/, MyIA.AI.Notebooks/SymbolicAI/SMT/, MyIA.AI.Notebooks/SymbolicAI/SemanticWeb/, MyIA.AI.Notebooks/SymbolicAI/SmartContracts/, MyIA.AI.Notebooks/SymbolicAI/SymbolicLearning/, MyIA.AI.Notebooks/SymbolicAI/Tweety/, MyIA.AI.Notebooks/cross-series/).
+    # 1504 notebooks (sous-arbres: MyIA.AI.Notebooks/CaseStudies/, MyIA.AI.Notebooks/Complexity/, MyIA.AI.Notebooks/Compression/, MyIA.AI.Notebooks/GameTheory/, MyIA.AI.Notebooks/GenAI/00-GenAI-Environment/, MyIA.AI.Notebooks/GenAI/Audio/, MyIA.AI.Notebooks/GenAI/CaseStudies/, MyIA.AI.Notebooks/GenAI/FallacyDetection/, MyIA.AI.Notebooks/GenAI/FineTuning/, MyIA.AI.Notebooks/GenAI/Image/, MyIA.AI.Notebooks/GenAI/Integrations-DotNet/, MyIA.AI.Notebooks/GenAI/Integrations-DotNet/Aspire/, MyIA.AI.Notebooks/GenAI/Plateformes-Conversationnelles/, MyIA.AI.Notebooks/GenAI/PostTraining/, MyIA.AI.Notebooks/GenAI/RAG-et-Memoire-Semantique/, MyIA.AI.Notebooks/GenAI/Security/, MyIA.AI.Notebooks/GenAI/SemanticKernel/, MyIA.AI.Notebooks/GenAI/Texte/, MyIA.AI.Notebooks/GenAI/Vibe-Coding/, MyIA.AI.Notebooks/GenAI/Video/, MyIA.AI.Notebooks/IIT/, MyIA.AI.Notebooks/ML/, MyIA.AI.Notebooks/NLP/, MyIA.AI.Notebooks/Probas/, MyIA.AI.Notebooks/QuantConnect/, MyIA.AI.Notebooks/RL/, MyIA.AI.Notebooks/Search/, MyIA.AI.Notebooks/Sudoku/, MyIA.AI.Notebooks/SymbolicAI/, MyIA.AI.Notebooks/SymbolicAI/Argument_Analysis/, MyIA.AI.Notebooks/SymbolicAI/Lean/, MyIA.AI.Notebooks/SymbolicAI/Planners/, MyIA.AI.Notebooks/SymbolicAI/SMT/, MyIA.AI.Notebooks/SymbolicAI/SemanticWeb/, MyIA.AI.Notebooks/SymbolicAI/SmartContracts/, MyIA.AI.Notebooks/SymbolicAI/SymbolicLearning/, MyIA.AI.Notebooks/SymbolicAI/Tweety/, MyIA.AI.Notebooks/cross-series/).
     - "MyIA.AI.Notebooks/CaseStudies/Diagnostic-Medical/solution/Diagnostic-Medical.ipynb"
     - "MyIA.AI.Notebooks/CaseStudies/Diagnostic-Medical/student/Diagnostic-Medical.ipynb"
     - "MyIA.AI.Notebooks/CaseStudies/Oncology-Planning/solution/Oncology-Planning.ipynb"
@@ -728,6 +740,7 @@ project:
     - "MyIA.AI.Notebooks/Complexity/Complexity-04d-Secretaire-Matroidal-Python.ipynb"
     - "MyIA.AI.Notebooks/Complexity/Complexity-05-CountingHarder-Permanent.ipynb"
     - "MyIA.AI.Notebooks/Complexity/Complexity-05b-AaronsonArkhipov-PermanenteBosonSampling.ipynb"
+    - "MyIA.AI.Notebooks/Complexity/Complexity-06-Simuler-Circuit-Quantique-Classiquement-Python.ipynb"
     - "MyIA.AI.Notebooks/Complexity/Complexity-06b-AaronsonGottesman-Dequantification-Python.ipynb"
     - "MyIA.AI.Notebooks/Complexity/Complexity-07-KolmogorovBornee-Sequences-Python.ipynb"
     - "MyIA.AI.Notebooks/Compression/Compression-01-ShannonFano-Prefixe-Python.ipynb"
@@ -849,6 +862,7 @@ project:
     - "MyIA.AI.Notebooks/GenAI/00-GenAI-Environment/00-4-Environment-Validation.ipynb"
     - "MyIA.AI.Notebooks/GenAI/00-GenAI-Environment/00-5-ComfyUI-Local-Test.ipynb"
     - "MyIA.AI.Notebooks/GenAI/00-GenAI-Environment/00-6-Local-Docker-Deployment.ipynb"
+    - "MyIA.AI.Notebooks/GenAI/00-GenAI-Environment/00-7-Terminal-Long-Runs.ipynb"
     - "MyIA.AI.Notebooks/GenAI/Audio/01-Foundation/01-1-OpenAI-TTS-Intro.ipynb"
     - "MyIA.AI.Notebooks/GenAI/Audio/01-Foundation/01-2-OpenAI-Whisper-STT.ipynb"
     - "MyIA.AI.Notebooks/GenAI/Audio/01-Foundation/01-3-Basic-Audio-Operations.ipynb"
@@ -1160,6 +1174,7 @@ project:
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-23-PersonaCatastrophe-Python.ipynb"
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-24-WorkspaceIgnition-Python.ipynb"
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-25-InoculationRL-Python.ipynb"
+    - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-25b-StratificationInterTailles-Python.ipynb"
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-26-SignalingConvention-Python.ipynb"
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-27-SymbolInvention-Python.ipynb"
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-28-CollectiveAdoption-Python.ipynb"
@@ -1187,6 +1202,7 @@ project:
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-44-GeometryOfTruth-Python.ipynb"
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-46-Strate7-FreeCoordinates-Python.ipynb"
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-47-PainAxisDistillation-Python.ipynb"
+    - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-47b-PainAxisInstrument-Python.ipynb"
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-Annexe-ProxyContextuality.ipynb"
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-Argumentation-BeliefTrajectories.ipynb"
     - "MyIA.AI.Notebooks/IIT/ICT-Series/ICT-Argumentation-QBFAcceptance.ipynb"
@@ -1348,6 +1364,7 @@ project:
     - "MyIA.AI.Notebooks/NLP/05-HMM-Viterbi.ipynb"
     - "MyIA.AI.Notebooks/Probas/Applications/Percolation/Percolation-01-Supercritique-Python.ipynb"
     - "MyIA.AI.Notebooks/Probas/Applications/Percolation/Percolation-02-Lean.ipynb"
+    - "MyIA.AI.Notebooks/Probas/Applications/Percolation/Percolation-03-Critique-Python.ipynb"
     - "MyIA.AI.Notebooks/Probas/Applications/Pyro_RSA_Hyperbole.ipynb"
     - "MyIA.AI.Notebooks/Probas/Applications/Quotients-Fibres-Recollement-Python.ipynb"
     - "MyIA.AI.Notebooks/Probas/DecisionTheory/Actuariat/Actuariat-01-Prime-Pure-Chargement.ipynb"
@@ -1355,6 +1372,7 @@ project:
     - "MyIA.AI.Notebooks/Probas/DecisionTheory/Actuariat/Actuariat-03-Actuarial-Credibility.ipynb"
     - "MyIA.AI.Notebooks/Probas/DecisionTheory/Actuariat/Actuariat-04-Ruine-Lundberg.ipynb"
     - "MyIA.AI.Notebooks/Probas/DecisionTheory/Actuariat/Actuariat-05-Valeur-Info-Souscription.ipynb"
+    - "MyIA.AI.Notebooks/Probas/DecisionTheory/Causal-Bridges/CausalBridges-00-PearlLadder-Intro-Python.ipynb"
     - "MyIA.AI.Notebooks/Probas/DecisionTheory/Causal-Bridges/CausalBridges-01-Do-Calculus.ipynb"
     - "MyIA.AI.Notebooks/Probas/DecisionTheory/Causal-Bridges/CausalBridges-02-Dowhy-Estimand-Intervention.ipynb"
     - "MyIA.AI.Notebooks/Probas/DecisionTheory/Causal-Bridges/CausalBridges-03-Dowhy-Contrefactuel-Individuel.ipynb"
@@ -1403,6 +1421,7 @@ project:
     - "MyIA.AI.Notebooks/Probas/Infer/Infer-7-Skills-IRT.ipynb"
     - "MyIA.AI.Notebooks/Probas/Infer/Infer-8-TrueSkill.ipynb"
     - "MyIA.AI.Notebooks/Probas/Infer/Infer-9-Classification.ipynb"
+    - "MyIA.AI.Notebooks/Probas/Probas-KLS-Concentration.ipynb"
     - "MyIA.AI.Notebooks/Probas/PyMC/PyMC-01-Setup.ipynb"
     - "MyIA.AI.Notebooks/Probas/PyMC/PyMC-02-Gaussian-Mixtures.ipynb"
     - "MyIA.AI.Notebooks/Probas/PyMC/PyMC-02b-Debugging-Python.ipynb"
@@ -1919,6 +1938,7 @@ project:
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/ANALYSE/ANALYSE-02-Tao-Lean-Python.ipynb"
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/ANALYSE/ANALYSE-03-PFR-Lean.ipynb"
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/ANALYSE/ANALYSE-04-PFR-Primitives-Python.ipynb"
+    - "MyIA.AI.Notebooks/SymbolicAI/Lean/ANALYSE/ANALYSE-08-Ramsey-VdW.ipynb"
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/ANALYSE/ANALYSE-09-Tuilage-Aperiodique.ipynb"
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/ANALYSE/ANALYSE-09-Tuilage-Aperiodique_en.ipynb"
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/Geometry/Geometry-01-From-Figure-To-Equation.ipynb"
@@ -1991,6 +2011,7 @@ project:
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/Lean-34b-FairBot-Loeb.ipynb"
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/Lean-36-Structures-Finies-MUH-Lean.ipynb"
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/Lean-37-Capstone-Serre100.ipynb"
+    - "MyIA.AI.Notebooks/SymbolicAI/Lean/Lean-38-Capstone-Geometry-Lean-Python.ipynb"
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/Serre100/01-corps-finis-borne-hasse.ipynb"
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/Serre100/02-valeurs-zeta-multiples-finies.ipynb"
     - "MyIA.AI.Notebooks/SymbolicAI/Lean/Serre100/03-cohomologie-cech-espaces-finis.ipynb"
diff --git a/docs/README.md b/docs/README.md
index c233e43921..e4b7e03f9c 100644
--- a/docs/README.md
+++ b/docs/README.md
@@ -279,6 +279,7 @@ Documents de recherche durables fondant les EPICs de R&D (grade A-recherche). Di
 | [research/cartier-miller-p1-elliptic-prefix.md](research/cartier-miller-p1-elliptic-prefix.md) | Lecture ligne-par-ligne de elliptic_prefix.py (P1 elliptic curve Cartier-Miller, EPIC #19452) — composante Schoof+BSGS quarter-point, stdlib only, 8 composants documentés. Suite #19487 (P0 cartography). |
 | [research/cartier-miller-p1-point-count.md](research/cartier-miller-p1-point-count.md) | Lecture ligne-par-ligne de `point_count.py` (P1 Schoof + BSGS, EPIC #19452) — stdlib only, 0 import, 9 composants documentés. Suite des PRs #19487 (P0 cartography), #19488 (P1 elliptic_prefix), #19493 (P1 pilot). |
 | [research/cartier-miller-p1-integration.md](research/cartier-miller-p1-integration.md) | Cross-check 4 backends de la P1 Cartier-Miller (EPIC #19452) : elliptic_prefix + pilot + point_count + C++ pseudocode, confrontation littérale aux `output.json` du pin reproductible `37a9b72`. P1+ à venir (P2 Lean Mathlib, P3+ multi-cycle) |
+| [research/percolation-03-critique-results.md](research/percolation-03-critique-results.md) | Résultats palier 3 (#19494, PR #19537) — physique du point critique de la percolation 2D (β/ν = 5/36 exact Kesten 1980, τ' = 187/91 Stauffer, p_c(L) → 1/2). Instrument `networkx`, sweep `L ∈ {8, 16, 32, 64}`. Verdict honnête : β/ν asymptote non tenue à L ≤ 64 (corrections d'échelle finie), p_c(L) monotone PASS conforme |
 | [research/slide-agents-marptoslidev-scoping.md](research/slide-agents-marptoslidev-scoping.md) | Scoping Marp→Slidev pour les agents de slides (c.1110, #19578) — inventaire firsthand (12 Marp coexistants avec 18 Slidev, configs et outils legacy), diagnostic des fronts communs (format Marp, PNG rendering mort-né, sk-agent vision périmé), 4 voies arbitrées (réécrire/fusionner/retirer/legacy), recommandation voie 1 (réécrire pour Slidev, 3 raisons mesurées). Critère de fermeture documenté ; arbitrage user ou coordinateur requis pour passer à la phase 2 |
 
 ## Audit sémantique cross-famille (docs/audit/)
diff --git a/docs/research/percolation-03-critique-results.md b/docs/research/percolation-03-critique-results.md
new file mode 100644
index 0000000000..4f898fff94
--- /dev/null
+++ b/docs/research/percolation-03-critique-results.md
@@ -0,0 +1,124 @@
+# Percolation-03 critique — résultats d'exécution (palier 3, c.1104)
+
+> **Suite directe de [#19494](https://github.com/jsboige/CoursIA/issues/19494)** (plan de croissance Percolation 03 critique / 04 sharpness ×2) et du scoping memo §8 (`percolation-03-critique-scoping.md`, c.1103, PR #19531) — exécution effective au pin `70ece81334` (main HEAD c.1104).
+
+## Sections (8)
+
+1. **Cible** — physique du point critique (β/ν, τ', p_c(L) → 1/2)
+2. **Architecture pipeline** — sweep `L × p × seeds`, instrument `networkx`
+3. **Mesures effectuées** — 580 runs, ~4 s wall-clock mono-thread
+5. **Verdict β/ν** — mesure vs cible asymptotique
+6. **Verdict τ'** — distribution de tailles
+7. **Verdict p_c(L)** — convergence monotone vers 1/2
+8. **Limitations et extensions** — corrections d'échelle finie, extensions L ∈ {128, 256}
+
+---
+
+## 1. Cible
+
+Mesurer la **physique du point critique** de la percolation 2D — exposants
+`β/ν = 5/36` (Kesten 1980, exact), `τ' = 187/91` (Stauffer computer-exp),
+convergence `p_c(L) → 1/2` quand `L → ∞`.
+
+**Ce qui distingue 03 de 01** : 01 mesure le *géant* en régime `p > p_c`
+(where il est stable). 03 mesure le *géant* en régime `p ≈ p_c` (where il
+**fluctue**) et la **distribution de tailles** des composantes connexes.
+
+## 3. Architecture pipeline
+
+- **L ∈ {8, 16, 32, 64}** (4 valeurs)
+- **p ∈ {0.45, 0.48, 0.50, 0.52, 0.55}** (5 valeurs)
+- **seeds = 29 par cellule `(L, p)`** : 5 canoniques (`{0,1,7,42,99}`) + 24 auxiliaires `range(100, 124)` (le scoping memo disait 32 seeds ; correction c.1104 : la liste effective est 29)
+- **Total : 4 × 5 × 29 = 580 simulations** (vs 640 visées)
+- `networkx.grid_2d_graph(L, L, periodic=True)` + `networkx.connected_components`
+
+## 4. Mesures effectuées
+
+- **Wall-clock total** : 3.8 s (640-budget vs 580 effectives) — soit ~0.007 s/sim en moyenne
+- **L=64 dominant** : ~17 ms/sim (vs L=8 ~0.3 ms/sim)
+- 0 erreur, aucune cellule `NotImplementedError` (C.1 conforme)
+- Outputs présents et cohérents (C.2 conforme)
+
+**Tableau des mesures `M(L, p) / L²` :**
+
+| L | p=0.45 | p=0.48 | p=0.50 | p=0.52 | p=0.55 |
+|---|--------|--------|--------|--------|--------|
+| 8  | 0.642 | 0.753 | 0.789 | 0.848 | 0.902 |
+| 16 | 0.493 | 0.661 | 0.732 | 0.802 | 0.872 |
+| 32 | 0.299 | 0.549 | 0.710 | 0.814 | 0.895 |
+| 64 | 0.118 | 0.393 | 0.629 | 0.793 | 0.881 |
+
+À `p = 0.50` (point critique), `M/L²` décroît lentement avec `L` (0.789 →
+0.732 → 0.710 → 0.629) — c'est la signature de la criticalité.
+
+## 5. Verdict β/ν
+
+**Régression log-log** sur `M(L, 1/2) / L²` vs `L` :
+- `β/ν mesuré = 0.102` (exact : 5/36 ≈ 0.139)
+- `R² = 0.946`
+- **Verdict** : `FAIL` — slope trop faible (~26% sous la valeur asymptotique)
+
+**Cause** : sur `L ∈ {8, 16, 32, 64}`, les corrections d'échelle finie
+dominent encore. La loi `M(L, 1/2) / L² ~ L^{-β/ν}` n'est asymptotique
+que pour `L ≫ ξ` (longueur de corrélation). À `p = 1/2`, `ξ = ∞`, donc
+l'asymptote n'est jamais vraiment atteinte — il faut `L` arbitrairement
+grand pour approcher `β/ν = 5/36`.
+
+## 6. Verdict τ'
+
+**Distribution de tailles** `P(s)` au point critique, fit log-log sur le
+bulk entre `s_min` et `s_max` pour chaque `L` :
+- `τ' moyen (sur L ∈ {8, 16, 32, 64}) = 2.20`
+- Exact : `187/91 ≈ 2.055`
+- **Verdict** : `FAIL` — `τ'` mesuré trop grand (~7% au-dessus de l'exact).
+
+**Cause** : le bulk à `L = 8` (taille totale 64) contient beaucoup de
+composantes triviales (`s = 1`, `s = 2`), ce qui rend la distribution
+**non asymptotique**. À `L = 64`, `τ'` est plus proche de l'exact mais
+l'effet cumulé sur 4 L tire la moyenne au-dessus.
+
+## 7. Verdict p_c(L) → 1/2
+
+**Interpolation linéaire** de `M(L, p) / L² = 1/2` :
+
+| L | p_c(L) mesuré |
+|---|---------------|
+| 8  | < 0.45 (giant fraction > 0.5 même à p=0.45 — hors plage) |
+| 16 | 0.4513 |
+| 32 | 0.4741 |
+| 64 | 0.4890 |
+
+**Convergence monotone** `0.4513 → 0.4741 → 0.4890 → 1/2` ✓
+
+C'est la **preuve directe** de la loi d'échelle finie : `p_c(L) = 1/2 + c · L^{-1/ν}`
+avec `1/ν = 3/4`. La convergence est conforme à la théorie.
+
+## 8. Limitations et extensions
+
+**Verdict global** : les **tendances** sont conformes (M/L² à p_c décroît,
+p_c(L) → 1/2 monotone) ; les **valeurs asymptotiques** (β/ν = 0.139,
+τ' = 2.055) ne sont **pas encore atteintes** sur `L ∈ {8..64}`.
+
+**Cause** : corrections d'échelle finie dominent à `L ≤ 64`. La
+percolation finie sur tore `T_n` ne reproduit l'asymptotique que pour
+`L ≫ ξ` — et à `p_c(ℤ²)`, `ξ = ∞`.
+
+**Extensions possibles** (à arbitrer pour un carnet ultérieur) :
+
+1. **Étendre le sweep à `L ∈ {128, 256, 512}`** — 5× la plage actuelle,
+   au prix de ~3 min supplémentaires par `(L, p)` × 32 seeds × 5 p.
+   Permettrait de mesurer `β/ν ∈ [0.12, 0.14]` (cible 0.139 ± 0.01).
+2. **Augmenter le nombre de seeds** à 64-256 par `(L, p)` pour
+   lisser les fluctuations à petit `L` (actuellement ±30% à L=8).
+3. **Migrer vers `scipy.sparse.csgraph.connected_components`** pour
+   ~10× speed-up et permettre L=1024 ou plus.
+
+**Convention honnête** : les valeurs rapportées ici sont les **mesures
+réelles** du sweep actuel, pas des ajustements cherry-picking. Un
+lecteur qui augmente `L` jusqu'à 512 doit voir `β/ν` converger vers
+0.139. Cette convergence est l'objet de la loi d'échelle finie, pas
+une coïncidence.
+
+---
+
+**Cycle c.1104, lane `myia-po-2023:CoursIA-2`** : exécution effective livrée. PR à publier.
\ No newline at end of file