Publish Ropedia Xperience-10M derived artifacts

EVIDENCE_CONTRACT.md

@@ -0,0 +1,28 @@
+# Evidence Contract
+
+This project is intentionally audit-first. Every visible claim should point to a
+local artifact that a reader can inspect before trusting the dashboard.
+
+| Claim | Current evidence | Status | Boundary |
+| --- | --- | --- | --- |
+| The public Xperience-10M sample has been converted into aligned model windows. | `results/episode_task_suite/windows.csv`, `results/episode_task_suite/shared_windows.npz`, `results/episode_task_suite/summary_report.json` | Verified for 5,821 frames and 1,161 windows | One public sample episode only |
+| The current feature contract is explicit and reviewable. | `results/episode_task_suite/feature_manifest.json`, `results/episode_task_suite/available_modalities.json` | Verified for an 8,378-d feature vector | Audio is present in MP4 streams but not yet a feature block |
+| The 12 task heads are real scripts and artifacts, not presentation placeholders. | `scripts/episode_task_suite.py`, `results/episode_task_suite/*/metrics.json`, `results/episode_task_suite/*/predictions.*` | Verified for all 12 task definitions | Chronological single-episode split, not cross-episode generalization |
+| Minimal and neural heads use the same task contracts. | `scripts/neural_task_models.py`, `results/episode_task_suite/neural_mlp/`, `docs/assets/task_architectures.png` | Verified for 12 minimal heads and 12 neural MLP heads | Small heads only; not a foundation model |
+| Four Ropedia research directions are mapped honestly as direct, proxy, or diagnostic evidence. | `results/episode_task_suite/research_directions/research_direction_taxonomy.json`, `docs/data/research_directions.json` | Verified taxonomy | Some directions remain proxy-only |
+| Four extra direction probes are coded and evaluated. | `results/episode_task_suite/research_direction_extensions/research_direction_extension_results.json`, `docs/data/research_direction_extensions.json` | Verified single-episode probes | Not full human modeling, neural rendering, intent modeling, or world modeling solutions |
+| Qwen3-Omni infrastructure has passed technical smoke checks. | `results/omni_finetune/RUN_REPORT.md`, `results/omni_finetune/dataset_manifest.json`, `results/omni_finetune/metrics_eval.json` | Smoke-only evidence | One episode, 128 train windows; not a 32-episode pilot |
+| The real 32-episode LoRA pilot is blocked on gated data access, not on repo presentation. | `results/omni_finetune/DATA_BLOCKER_REPORT.md`, `results/omni_finetune/A100_HF_RELAY_STATUS.md`, `results/omni_finetune/source_discovery.json` | Blocker documented | No 32-episode metric should be claimed until the gate passes |
+
+## Review Order
+
+1. Read `docs/assets/task_suite_infographic.png` for the high-level map.
+2. Inspect `results/episode_task_suite/summary_report.json` for the task and
+   metric source of truth.
+3. Inspect `results/episode_task_suite/feature_manifest.json` to see which
+   modalities enter the current feature vector.
+4. Inspect `results/episode_task_suite/neural_mlp/` to compare minimal and
+   neural heads under the same splits.
+5. Inspect `results/omni_finetune/DATA_BLOCKER_REPORT.md` before interpreting
+   any Qwen3-Omni artifact.
+

PROJECT_README.md

@@ -28,6 +28,23 @@ into:
 - metrics, predictions, model weights, manifests, charts, and a static website,
 - a clear explanation of what a single episode can and cannot prove.
 
+## Evidence Contract
+
+This repo is organized around an explicit proof boundary:
+
+| Claim layer | Evidence | Boundary |
+| --- | --- | --- |
+| Data windows | `results/episode_task_suite/windows.csv`, `shared_windows.npz`, `summary_report.json` | one public sample episode |
+| Feature contract | `results/episode_task_suite/feature_manifest.json`, `available_modalities.json` | 8,378 current features; audio documented but not featurized |
+| 12-task suite | `scripts/episode_task_suite.py`, per-task `metrics.json`, predictions | chronological single-episode split |
+| Neural heads | `scripts/neural_task_models.py`, `results/episode_task_suite/neural_mlp/` | compact MLP heads, not a foundation model |
+| Research directions | `research_direction_taxonomy.json`, extension probe results | direct/proxy/diagnostic evidence, not full solutions |
+| Qwen3-Omni | `results/omni_finetune/DATA_BLOCKER_REPORT.md`, `A100_HF_RELAY_STATUS.md` | smoke-only until 32 valid episodes are available |
+
+Read the full contract in [`EVIDENCE_CONTRACT.md`](EVIDENCE_CONTRACT.md), or
+consume the machine-readable copy at
+[`docs/data/evidence_contract.json`](docs/data/evidence_contract.json).
+
 ## Dataset Modality Coverage
 
 The Xperience-10M sample is a 4D multimodal episode source spanning video,
@@ -139,6 +156,7 @@ results/
 docs/
   index.html                        # GitHub Pages dashboard
   data/summary_metrics.json         # website-readable metrics bundle
+  data/evidence_contract.json       # machine-readable proof boundary
   data/research_directions.json     # four-track website data bundle
   data/research_direction_extensions.json # four extra probe data bundle
   data/task_walkthroughs.json       # beginner task explanation data bundle

README.md

@@ -44,6 +44,17 @@ real 32-episode held-out metrics.
 
 This is the reviewable half of the project. You can inspect the task outputs, compare the committed metrics, and understand the single-episode limitations without downloading the raw videos first.
 
+## Evidence Contract
+
+| Claim layer | Evidence | Boundary |
+| --- | --- | --- |
+| Data windows | `results/episode_task_suite/windows.csv`, `shared_windows.npz`, `summary_report.json` | one public sample episode |
+| Feature contract | `results/episode_task_suite/feature_manifest.json`, `available_modalities.json` | 8,378 current features; audio documented but not featurized |
+| 12-task suite | per-task `metrics.json`, predictions, confusion matrices | chronological single-episode split |
+| Neural heads | `results/episode_task_suite/neural_mlp/` | compact MLP heads, not a foundation model |
+| Research directions | `research_direction_taxonomy.json`, extension probe results | direct/proxy/diagnostic evidence, not full solutions |
+| Qwen3-Omni | `DATA_BLOCKER_REPORT.md`, `A100_HF_RELAY_STATUS.md` | smoke-only until 32 valid episodes are available |
+
 ## What Is Included
 
 - `results/**/*.json`: verified metrics and metadata for minimal and neural MLP runs
@@ -53,6 +64,7 @@ This is the reviewable half of the project. You can inspect the task outputs, co
 - `docs/assets/*.svg` and `docs/assets/*.png`: generated diagrams, charts, and ChatGPT-image-backed overview figures
 - `docs/assets/task_suite_infographic.png`: ChatGPT-image-backed infographic with low-resolution public-sample modality thumbnails, including audio waveform context, and verified metric overlays
 - `docs/data/summary_metrics.json`: dashboard-readable summary bundle
+- `docs/data/evidence_contract.json`: machine-readable proof boundary
 - `docs/data/research_directions.json`: generated four-track taxonomy for the website
 - `docs/data/research_direction_extensions.json`: four extra data-backed probes, one per research direction
 - `docs/data/task_walkthroughs.json`: beginner-oriented input/process/output guide for all 12 tasks

docs/data/evidence_contract.json

@@ -0,0 +1,91 @@
+{
+  "project": "Ropedia Xperience-10M Task Suite",
+  "scope": "single public Xperience-10M sample episode",
+  "claims": [
+    {
+      "id": "aligned_windows",
+      "claim": "The public Xperience-10M sample has been converted into aligned model windows.",
+      "status": "verified",
+      "evidence": [
+        "results/episode_task_suite/windows.csv",
+        "results/episode_task_suite/shared_windows.npz",
+        "results/episode_task_suite/summary_report.json"
+      ],
+      "boundary": "5,821 frames, 1,161 windows, one public sample episode"
+    },
+    {
+      "id": "feature_contract",
+      "claim": "The current feature contract is explicit and reviewable.",
+      "status": "verified",
+      "evidence": [
+        "results/episode_task_suite/feature_manifest.json",
+        "results/episode_task_suite/available_modalities.json"
+      ],
+      "boundary": "8,378-d feature vector; audio is documented but not yet featurized"
+    },
+    {
+      "id": "twelve_tasks",
+      "claim": "The 12 task heads are real scripts and artifacts, not presentation placeholders.",
+      "status": "verified",
+      "evidence": [
+        "scripts/episode_task_suite.py",
+        "results/episode_task_suite/*/metrics.json",
+        "results/episode_task_suite/*/predictions.*"
+      ],
+      "boundary": "chronological single-episode split, not cross-episode generalization"
+    },
+    {
+      "id": "minimal_vs_neural",
+      "claim": "Minimal and neural heads use the same task contracts.",
+      "status": "verified",
+      "evidence": [
+        "scripts/neural_task_models.py",
+        "results/episode_task_suite/neural_mlp/",
+        "docs/assets/task_architectures.png"
+      ],
+      "boundary": "small heads only; not a foundation model"
+    },
+    {
+      "id": "research_directions",
+      "claim": "Four Ropedia research directions are mapped honestly as direct, proxy, or diagnostic evidence.",
+      "status": "verified",
+      "evidence": [
+        "results/episode_task_suite/research_directions/research_direction_taxonomy.json",
+        "docs/data/research_directions.json"
+      ],
+      "boundary": "some directions remain proxy-only"
+    },
+    {
+      "id": "direction_extensions",
+      "claim": "Four extra direction probes are coded and evaluated.",
+      "status": "verified",
+      "evidence": [
+        "results/episode_task_suite/research_direction_extensions/research_direction_extension_results.json",
+        "docs/data/research_direction_extensions.json"
+      ],
+      "boundary": "single-episode probes, not full research-direction solutions"
+    },
+    {
+      "id": "qwen3_omni_smoke",
+      "claim": "Qwen3-Omni infrastructure has passed technical smoke checks.",
+      "status": "smoke_only",
+      "evidence": [
+        "results/omni_finetune/RUN_REPORT.md",
+        "results/omni_finetune/dataset_manifest.json",
+        "results/omni_finetune/metrics_eval.json"
+      ],
+      "boundary": "one episode, 128 train windows; not a 32-episode pilot"
+    },
+    {
+      "id": "thirty_two_episode_gate",
+      "claim": "The real 32-episode LoRA pilot is blocked on gated data access, not on repo presentation.",
+      "status": "blocked_by_data_access",
+      "evidence": [
+        "results/omni_finetune/DATA_BLOCKER_REPORT.md",
+        "results/omni_finetune/A100_HF_RELAY_STATUS.md",
+        "results/omni_finetune/source_discovery.json"
+      ],
+      "boundary": "no 32-episode metric should be claimed until the gate passes"
+    }
+  ]
+}

docs/index.html

@@ -341,6 +341,44 @@
     .callout h3, .artifact h3 { margin: 0 0 8px; font-size: 17px; }
     .callout p, .artifact p { margin: 0; color: #aab5a5; line-height: 1.6; }
 
+    .evidence-grid {
+      display: grid;
+      grid-template-columns: repeat(2, minmax(0, 1fr));
+      gap: 18px;
+    }
+    .evidence-card {
+      border: 1px solid var(--line);
+      border-radius: var(--radius);
+      padding: 22px;
+      background:
+        linear-gradient(180deg, rgba(164, 242, 127, 0.06), rgba(7, 18, 7, 0.88)),
+        var(--surface);
+      min-height: 230px;
+      display: grid;
+      gap: 12px;
+      align-content: start;
+    }
+    .evidence-card h3 { margin: 0; font-size: 18px; line-height: 1.2; }
+    .evidence-card p { margin: 0; color: var(--muted); line-height: 1.6; }
+    .evidence-card code { color: var(--ink); font-family: var(--font-mono); font-size: 12px; }
+    .evidence-links {
+      display: flex;
+      flex-wrap: wrap;
+      gap: 10px;
+      margin-top: 4px;
+    }
+    .evidence-links a {
+      border: 1px solid var(--soft-line);
+      border-radius: 6px;
+      color: var(--blue);
+      font-size: 13px;
+      font-weight: 740;
+      padding: 7px 9px;
+      text-decoration: none;
+      background: rgba(2, 5, 2, 0.42);
+    }
+    .evidence-links a:hover { border-color: var(--green); color: var(--ink); }
+
     .models {
       display: grid;
       grid-template-columns: repeat(2, minmax(0, 1fr));
@@ -576,7 +614,7 @@
     @media (max-width: 960px) {
       .hero-inner, .two-col { grid-template-columns: 1fr; }
       .hero-inner { min-height: 0; }
-      .hero-stats, .models, .task-grid, .artifact-grid, .callout-row, .direction-grid, .baseline-strip, .extension-grid, .walkthrough-grid { grid-template-columns: repeat(2, minmax(0, 1fr)); }
+      .hero-stats, .models, .task-grid, .artifact-grid, .evidence-grid, .callout-row, .direction-grid, .baseline-strip, .extension-grid, .walkthrough-grid { grid-template-columns: repeat(2, minmax(0, 1fr)); }
       .chart-grid { grid-template-columns: 1fr; }
       .section-head { display: block; }
       .section-head p { margin-top: 14px; }
@@ -584,7 +622,7 @@
     }
     @media (max-width: 640px) {
       .wrap { width: min(100% - 28px, var(--max)); }
-      .hero-stats, .models, .task-grid, .artifact-grid, .chart-grid, .callout-row, .direction-grid, .baseline-strip, .extension-grid, .walkthrough-grid, .walk-flow { grid-template-columns: 1fr; }
+      .hero-stats, .models, .task-grid, .artifact-grid, .evidence-grid, .chart-grid, .callout-row, .direction-grid, .baseline-strip, .extension-grid, .walkthrough-grid, .walk-flow { grid-template-columns: 1fr; }
       .hero-inner, section { padding: 46px 0; }
       .signal { grid-template-columns: 1fr; }
       .signal strong { text-align: left; }
@@ -609,8 +647,8 @@
         <span>Ropedia Xperience-10M Task Suite</span>
       </a>
       <div class="nav-links" aria-label="Page navigation">
+        <a href="#evidence">Evidence</a>
         <a href="#suite">Suite</a>
-        <a href="#pipeline">Pipeline</a>
         <a href="#models">Models</a>
         <a href="#directions">Directions</a>
         <a href="#walkthroughs">Guide</a>
@@ -661,6 +699,53 @@
   </header>
 
   <main id="main">
+    <section id="evidence">
+      <div class="wrap">
+        <div class="section-head">
+          <h2>Evidence first, claims second.</h2>
+          <p>A top-level project should make its proof boundary visible. This ledger separates verified single-episode artifacts from smoke-only Qwen3-Omni work and the pending 32-episode gate.</p>
+        </div>
+        <div class="evidence-grid">
+          <article class="evidence-card">
+            <span class="status-pill">verified</span>
+            <h3>Aligned Xperience-10M sample windows</h3>
+            <p>5,821 frames become 1,161 synchronized 20-frame windows with an explicit 8,378-d feature contract.</p>
+            <div class="evidence-links">
+              <a href="https://github.com/ChaoYue0307/ropedia-xperience-10m-task-suite/blob/main/results/episode_task_suite/summary_report.json">summary_report.json</a>
+              <a href="https://github.com/ChaoYue0307/ropedia-xperience-10m-task-suite/blob/main/results/episode_task_suite/feature_manifest.json">feature_manifest.json</a>
+            </div>
+          </article>
+          <article class="evidence-card">
+            <span class="status-pill">verified</span>
+            <h3>12 minimal heads + 12 neural MLP heads</h3>
+            <p>Every task has a minimal interpretable head and a matching neural MLP run over the same windows, splits, and task contract.</p>
+            <div class="evidence-links">
+              <a href="https://github.com/ChaoYue0307/ropedia-xperience-10m-task-suite/tree/main/results/episode_task_suite">task artifacts</a>
+              <a href="https://github.com/ChaoYue0307/ropedia-xperience-10m-task-suite/tree/main/results/episode_task_suite/neural_mlp">neural_mlp/</a>
+            </div>
+          </article>
+          <article class="evidence-card">
+            <span class="status-pill">verified</span>
+            <h3>Four research directions are mapped honestly</h3>
+            <p>The Ropedia directions are labeled as direct, proxy, or diagnostic evidence, plus one coded extension probe per direction.</p>
+            <div class="evidence-links">
+              <a href="data/research_directions.json">directions JSON</a>
+              <a href="data/research_direction_extensions.json">extensions JSON</a>
+            </div>
+          </article>
+          <article class="evidence-card">
+            <span class="status-pill">blocked by data access</span>
+            <h3>Qwen3-Omni is smoke-only until 32 episodes land</h3>
+            <p>The H20/A100 pipeline exists, but current Qwen3-Omni artifacts use one episode and 128 train windows. No 32-episode metric is claimed.</p>
+            <div class="evidence-links">
+              <a href="https://github.com/ChaoYue0307/ropedia-xperience-10m-task-suite/blob/main/EVIDENCE_CONTRACT.md">evidence contract</a>
+              <a href="data/evidence_contract.json">machine JSON</a>
+            </div>
+          </article>
+        </div>
+      </div>
+    </section>
+
     <section id="suite">
       <div class="wrap">
         <div class="section-head">