File size: 12,219 Bytes
e647650 f7f39ba c325020 f7f39ba c325020 f7f39ba cf07180 b7a466b 4602161 b7a466b 4602161 f7f39ba 29331c9 cca436c 94a5118 cca436c 29331c9 c325020 6a1869c 29331c9 c325020 29331c9 e647650 cca436c c325020 cca436c 45c1706 cca436c cf07180 cca436c 9d58132 3e04138 c325020 3e04138 4173e02 c325020 4173e02 cf07180 4173e02 9d58132 c325020 9d58132 2c5b88c 756e790 2c5b88c e647650 04c0bde e647650 a07660e e647650 a07660e e647650 a07660e e647650 a07660e e647650 a07660e e647650 a07660e 4b0d658 f590d7e cf07180 45c1706 f590d7e 45c1706 f590d7e 0f9a8e2 03b872c 0f9a8e2 4b0d658 cf07180 4b0d658 04c0bde 4b0d658 7977885 149cadc 08a4bf0 c325020 08a4bf0 cf07180 08a4bf0 a6472b6 4b0d658 c325020 4b0d658 c325020 4bd6e11 4b0d658 cf07180 4b0d658 c325020 4b0d658 4bd6e11 4b0d658 7faed79 c325020 7faed79 a07660e e647650 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 | {
"project": "Ropedia Xperience-10M Task Suite",
"scope": "single public Xperience-10M sample episode",
"claims": [
{
"id": "project_status",
"claim": "A first-pass reader has a compact current-state summary.",
"status": "verified",
"evidence": [
"PROJECT_STATUS.md",
"docs/data/project_status.json"
],
"boundary": "summarizes existing evidence and current limitations"
},
{
"id": "research_roadmap",
"claim": "The research roadmap is explicit.",
"status": "current",
"evidence": [
"RESEARCH_ROADMAP.md",
"docs/data/research_roadmap.json"
],
"boundary": "connects public-sample task development to multi-episode data preparation, Qwen3-Omni LoRA, robustness runs, and larger omni-model extensions"
},
{
"id": "official_dataset_card_alignment",
"claim": "The public dataset description is aligned with the official gated Xperience-10M dataset card and public sample card.",
"status": "verified",
"evidence": [
"XPERIENCE10M_DATASET_CARD_ALIGNMENT.md",
"docs/data/xperience10m_dataset_card_alignment.json",
"https://huggingface.co/datasets/ropedia-ai/xperience-10m"
],
"boundary": "summarizes upstream public metadata, API listing facts, sample license/tooling, and dataset-card facts; does not grant access or mirror raw data"
},
{
"id": "source_alignment",
"claim": "Source facts, sample details, API-listing notes, and project coverage are validated across repo, website, and HF cards.",
"status": "verified",
"evidence": [
"SOURCE_ALIGNMENT_AUDIT.md",
"docs/data/source_alignment_audit.json",
"scripts/validate_source_alignment.py"
],
"boundary": "offline committed-fact check; does not fetch private gated data"
},
{
"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 inspectable.",
"status": "verified",
"evidence": [
"results/episode_task_suite/feature_manifest.json",
"results/episode_task_suite/available_modalities.json"
],
"boundary": "8,546-dimensional aligned multimodal window representation"
},
{
"id": "evaluation_protocol",
"claim": "The task evaluation protocol is explicit and generated from committed metrics.",
"status": "verified",
"evidence": [
"EVALUATION_PROTOCOL.md",
"docs/data/evaluation_protocol.json",
"scripts/build_evaluation_protocol.py"
],
"boundary": "defines windows, split, per-task metrics, leakage controls, and current limitations"
},
{
"id": "modality_atlas",
"claim": "The public sample modalities are inspectable without raw data redistribution.",
"status": "verified",
"evidence": [
"docs/data/modality_atlas.json",
"docs/assets/modalities/",
"docs/index.html"
],
"boundary": "derived thumbnails for presentation; raw data remains excluded"
},
{
"id": "task_surface_integrity",
"claim": "Public task cards stay readable for non-expert readers.",
"status": "verified",
"evidence": [
"docs/data/task_surface_integrity.json",
"scripts/validate_task_surface.py",
"docs/index.html"
],
"boundary": "presentation integrity for the public task surface"
},
{
"id": "figure_index",
"claim": "Public figures, charts, and modality thumbnails are indexed as project evidence.",
"status": "verified",
"evidence": [
"FIGURE_INDEX.md",
"docs/data/figure_index.json",
"scripts/build_figure_index.py"
],
"boundary": "records derived visual assets, dimensions, hashes, roles, and source scripts; raw Xperience-10M data remains excluded"
},
{
"id": "brand_assets",
"claim": "A project logo is consistently applied across public surfaces.",
"status": "verified",
"evidence": [
"docs/assets/brand/",
"docs/data/brand_assets.json",
"scripts/build_brand_assets.py"
],
"boundary": "generated logo and deterministic derivatives only; no raw dataset data or model weights"
},
{
"id": "twelve_tasks",
"claim": "The 12 task heads are implemented as scripts with saved metrics and predictions.",
"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_diagnostic_pilot",
"claim": "Qwen3-Omni has a verified selected-episode held-out diagnostic pilot.",
"status": "verified_diagnostic",
"evidence": [
"docs/data/omni_finetune_verified_result.json",
"results/omni_finetune/verified_public/",
"scripts/omni/package_verified_omni_result.py",
"scripts/omni/audit_verified_omni_package.py"
],
"boundary": "the pipeline is verified, but model quality is weak: JSON validity is below target and action/subtask metrics are low"
},
{
"id": "multi_episode_quality_improvement",
"claim": "The next Qwen3-Omni step is structured-output and task-quality improvement on the same selected split.",
"status": "active_next_step",
"evidence": [
"scripts/omni/run_128_fullsplit_parallel_export_8gpu.sh",
"docs/data/omni_finetune_verified_result.json",
"FOUNDATION_MODEL_PLAN.md"
],
"boundary": "stronger model quality requires output-format improvements and action/subtask error analysis"
},
{
"id": "scale_up_status_check",
"claim": "Older pilot path strings are tracked as setup-file provenance.",
"status": "verified",
"evidence": [
"scripts/validate_scope_claims.py",
"docs/data/scope_claims_audit.json"
],
"boundary": "run/path identifiers stay separate from completed held-out-episode results"
},
{
"id": "mirror_parity",
"claim": "Prepared GitHub and Hugging Face mirrors carry matching critical data, visual, HTML, and validator files.",
"status": "verified",
"evidence": [
"scripts/validate_mirror_parity.py",
"docs/data/mirror_parity.json"
],
"boundary": "compares prepared local mirror bundles before upload; live URLs are checked after publishing"
},
{
"id": "publication_package",
"claim": "The public GitHub and Hugging Face bundles contain the intended release files.",
"status": "verified",
"evidence": [
"scripts/validate_publication_package.py",
"docs/data/publication_audit.json"
],
"boundary": "checks public files, HF bundles, and current public-card assets; temporary local outputs are excluded"
},
{
"id": "website_integrity",
"claim": "The public website has checked local references.",
"status": "verified",
"evidence": [
"scripts/validate_website_integrity.py",
"docs/data/website_integrity.json"
],
"boundary": "checks local links, anchors, JSON data, and referenced images; external URLs are not fetched"
},
{
"id": "rendered_site_check",
"claim": "The rendered website walkthrough has a browser-level interaction check.",
"status": "verified",
"evidence": [
"RENDERED_SITE_CHECK.md",
"scripts/build_rendered_site_check.py",
"docs/data/rendered_site_check.json"
],
"boundary": "checks local page load, tab switch, walkthrough deep link, player controls, and console health"
},
{
"id": "quality_gates",
"claim": "The release gate is explicit.",
"status": "verified",
"evidence": [
"QUALITY_GATES.md",
"scripts/build_quality_gates.py",
"docs/data/quality_gates.json"
],
"boundary": "summarizes packaging and live-mirror checks; cross-episode model quality is measured by later held-out reports"
},
{
"id": "live_publication_status",
"claim": "The live public mirrors are checked after upload.",
"status": "verified",
"evidence": [
"scripts/verify_live_publication.py",
"docs/data/live_publication_status.json"
],
"boundary": "fetches public GitHub/HF URLs; it does not validate private training state"
},
{
"id": "citation_metadata",
"claim": "The project is externally citable and machine-readable.",
"status": "verified",
"evidence": [
"CITATION.cff",
"codemeta.json",
"docs/data/project_manifest.json",
"LICENSE"
],
"boundary": "code license does not override original Xperience-10M dataset terms"
},
{
"id": "project_path",
"claim": "A first-time reader has an explicit project path.",
"status": "verified",
"evidence": [
"docs/data/project_packet.json",
"ARTIFACT_GUIDE.md",
"docs/data/artifact_index.json",
"README.md",
"docs/index.html"
],
"boundary": "guides inspection across data, tasks, results, and scale-up status"
},
{
"id": "artifact_index",
"claim": "The core project artifacts are grouped for human reading and indexed with existence, size, and hash metadata where stable.",
"status": "verified",
"evidence": [
"ARTIFACT_GUIDE.md",
"scripts/build_artifact_index.py",
"docs/data/artifact_index.json"
],
"boundary": "selective source-of-truth catalog, not a complete inventory of every output file"
},
{
"id": "reproducibility_contract",
"claim": "The public reproduction path is documented with commands, expected outputs, and exact-match reproduction evidence.",
"status": "verified",
"evidence": [
"REPRODUCIBILITY.md",
"docs/data/reproducibility_matrix.json",
"notes/reproducibility_audit.md"
],
"boundary": "publicly reproduces the single-episode pipeline; multi-episode Qwen3-Omni metrics are added only after data preparation and held-out evaluation"
}
]
}
|