-
-
-
- 
-
- A multilingual public research surface for Xperience-10M: sample data, 20 embodied-AI tasks, baselines, Qwen3-Omni and Cosmos3 diagnostics, and foundation-model training directions. -
- - -- English · - 中文 · - Español · - Français · - Deutsch · - 日本語 · - 한국어 · - Português -
- - -
- 
| Signal | -Current public state | -
|---|---|
| Project identity |
- The same project logo mark is used across the GitHub README, GitHub Pages dashboard, Hugging Face Space, artifact dataset, model mirrors, favicon, and social preview. Ropedia is credited as the Xperience-10M data provider and releaser; this repository is the task-suite and evaluation layer built on that dataset. Reusable assets: logo mark and social card. | -
| Two-line contract | -Line 1: 1 sample episode for task construction and reproducibility. Line 2: 128 selected episodes for same-split metadata/raw baselines, Qwen3-Omni v6, and Cosmos3 diagnostics. | -
| Data explorer analysis | -A generated analysis layer separates the public sample, selected-128 feature exports, and authenticated Hugging Face gated full-dataset metadata with scope stats, split counts, modality breakdowns, and chart assets. | -
| 180 method-task records | -9 methods x 20 tasks = 180/180 scored records. The ledger separates 174 direct scores from 6 compact-proxy scores. | -
| 20 task contracts | -Action, procedure, transition, trajectory, contact, objects, language, retrieval, reconstruction, order, sync, long-horizon forecasting, interaction text, action-object binding, sensor bridging, camera sync, and transition timing. | -
| 4 research directions | -Human Modeling & Motion Understanding; 3D/4D Reconstruction & Neural Rendering; Egocentric Vision & Interaction; Scene Reconstruction & World Modeling. These are analysis groups over the same 20 tasks, not separate benchmark tiers. | -
| Line 1 methods | -Minimal and Neural MLP baselines cover all 20 tasks on the one public sample episode: 40/40 direct scores. | -
| Line 2 methods | -Metadata simple/NN, raw-feature simple/NN, Qwen3-Omni v6 LoRA, Cosmos3-Super Reasoner, and Cosmos3-Nano Future Window cover all 20 selected-128 task axes: 140/140 scores. | -
| 3 foundation pipelines | -Spatial intelligence, human-video world modeling, and vision-language-action pipelines are documented as training recipes with task mappings, input-output contracts, and model-evidence requirements. | -
| 1 unified target | -The long-term embodied foundation-model target connects perception, 3D memory, language-grounded reasoning, action, and planning without adding a new score axis. | -
| Public mirrors | -GitHub, GitHub Pages, HF Space, HF artifact dataset, HF baseline model repo, Qwen3-Omni and Cosmos3 model repos, and HF collection. | -
- 
-
- 
+
| Line | -Data unit | -Score statement | -Best use | -Read separately from | -
|---|---|---|---|---|
| 1 sample episode | -One public Xperience-10M sample episode: 5,821 frames, 1,161 aligned 20-frame windows, 8,546 feature dimensions. | -40/40 direct scores from Minimal and Neural MLP heads. | -Inspect the raw sample, understand file organization, reproduce the 20 task targets, and compare Minimal vs Neural MLP behavior inside one episode. | -The selected-128 comparison rows and any broader held-out model behavior. | -
| 128 selected episodes | -Selected held-out 96/16/16 split: 34,269 exported windows with public-safe processed features linked to official gated episode paths. The Hugging Face artifact dataset exposes these rows separately as selected_128_windows/selected_128; it is not mixed with the one-sample episode_sample/public_sample viewer. |
- 140/140 selected-128 scores: 134 direct + 6 compact-proxy. | -Compare same-split metadata/raw baselines, Qwen3-Omni v6, Cosmos3-Super, and Cosmos3-Nano while keeping the 6 compact-proxy cells visible. | -Direct raw-target measurements for the proxy-marked cells. | -
| Line | -Methods | -Tasks | -Scored records | -Direct scores | -Proxy scores | -
|---|---|---|---|---|---|
| 1 sample episode | -2 | -20 | -40/40 | -40 | -0 | -
| 128 selected episodes | -7 | -20 | -140/140 | -134 | -6 compact-proxy scores, each source-linked and reasoned. | -
| Total public matrix | -9 | -20 | -180/180 | -174 | -6 | -
| Evidence line | -Method block | -Methods | -Score statement | -Read as | -
|---|---|---|---|---|
| 1 sample episode | -Task-head baselines | -Minimal; Neural MLP | -40/40 direct scores. | -Task-lab reproducibility and simple-vs-neural behavior. | -
| 128 selected episodes | -Aligned baseline heads | -Metadata simple/NN; raw-feature simple/NN | -80/80 scores: 74 direct + 6 compact-proxy. | -Same-split metadata/raw-feature baseline comparison. | -
| 128 selected episodes | -Qwen3-Omni series | -Qwen3-Omni v6 LoRA | -20/20 direct scores from verified selected-128 Qwen3-Omni LoRA and task-specific probes. | -Trainable Qwen3-Omni diagnostic baseline on the selected-128 surface. | -
| 128 selected episodes | -Cosmos3 series | -Cosmos3-Super Reasoner; Cosmos3-Nano Future Window | -40/40 direct scores from verified public-safe reasoner and future-window artifacts. | -Cosmos3 reasoner and future-window diagnostics on the selected-128 surface. | -
| Run | -Purpose | -Main change | -Eval signal | -Use now | -
|---|---|---|---|---|
| v1 | Prove the selected-128 LoRA/eval/package loop. | First verified 96/16/16 selected-episode Qwen3-Omni LoRA run. | 448 eval; JSON 0.8750; contact 0.6451. | Lineage only. |
| v2 | Make answers schema-checked. | Structured-JSON contract with full-8-GPU LoRA on the same split. | 448 eval; JSON 0.9978; contact 0.7188. | Structured-output ablation. |
| v3 | Separate prompt/eval effects from training. | Strict-label prompt/eval over the v2 adapter; no new adapter training. | 448 eval; JSON 1.0000; contact 0.7210. | Prompt/eval ablation. |
| v4 | Test longer structured-JSON LoRA training. | New four-epoch full-8-GPU adapter on the same selected split. | 448 eval; JSON 1.0000; contact 0.7299. | Overfit/metric-tradeoff evidence. |
| v5 | Move to denser multiscale evaluation. | Multiscale cap96 export with 4,032 held-out predictions. | 4,032 eval; JSON 1.0000; contact 0.7865. | Pinned prior release; stronger on several non-contact metrics. |
| v6 | Publish the current Qwen 20-task row. | Rank64/lr5e-5 multiscale LoRA plus verified task-specific probes. | 4,032 eval; JSON 0.9990; contact 0.8177. | Current public 20-task Qwen3-Omni row. |
| Goal | -Start here | -Then inspect | -
|---|---|---|
| Understand quickly | -Project brief Project status |
- Dashboard | -
| Choose the published mirror | -Public evidence map | -evidence-map data | -
| Decode project terms | -Glossary | -glossary data | -
| Inspect the 20 tasks | -20-task guide | -task contract data task walkthroughs |
-
| Explore data scales | -Data explorer analysis | -website analysis section structured analysis record |
-
| Compare results | -Research takeaways | -two-line result summary 180-record result table radar data score/proxy audit |
-
| Understand one sample | -Single-episode explorer | -sample-file map feature manifest |
-
| Read foundation directions | -Three foundation pipelines | -pipeline contract data foundation model plan |
-
| Reproduce or audit | -Reproducibility Evidence contract |
- quality gates publication audit mirror parity |
-
| Capability | -What this project shows | -
|---|---|
| Multimodal data understanding | Parses the public sample into synchronized windows across video, audio, depth, pose/SLAM, mocap, IMU, calibration, and language-derived signals. |
| Task design | Defines 20 human-readable tasks in one unified public-sample suite, plus four direction-extension probes with inputs, outputs, process modules, metrics, and case-study walkthroughs. |
| Model and evaluation discipline | Runs minimal and compact neural baselines, records predictions/metrics, keeps chronological split boundaries explicit, and separates the sample readout from held-out comparison rows. |
| Scale-up planning | Connects the public-sample pipeline to 32/128-episode held-out pilots, Qwen3-Omni LoRA, Cosmos-style world-model tracks, policy/VLA tracks, and the future Xperience-native foundation-model pretraining goal. |
| Goal | -Best entry point | -
|---|---|
| Choose the right published mirror | Public evidence map evidence-map data |
| Resolve confusing terms and abbreviations | Glossary glossary data |
| Understand the whole project quickly | Project brief |
| See the visual research dashboard | GitHub Pages dashboard |
| Navigate the unified 20 tasks, four tracks, and scale-up plan | Interactive research roadmap 20-task suite note task contract data interactive roadmap data |
| Compare current task metrics | Research takeaways summary metrics |
| Compare possible foundation backbones | Foundation-model plan foundation-model data |
| Understand the future native pretraining goal | Native pretraining plan |
| See additional concrete project directions | Additional development directions direction data |
| Understand one model input | feature manifest window table |
| Check multi-episode data status | multi-episode data status |
| Surface | -What it is for | -
|---|---|
| GitHub repo | Source of truth for docs, scripts, generated data, validators, and commit history. |
| GitHub Pages dashboard | Best visual overview of the sample, 20 tasks, radar results, foundation directions, and resources. |
| Hugging Face Space | Hub-hosted copy of the dashboard and static app assets. |
| HF artifact dataset | Public-safe metrics, reports, website data, result packages, and derived evidence files. |
| HF baseline model repo | Minimal/neural baseline weights, figures, metrics, and mirrored task artifacts. |
| Qwen3-Omni and Cosmos3 model repos | Adapter-specific public weights or package cards when Qwen3-Omni v6, Cosmos3-Super, or Cosmos3-Nano runs are verified and publishable. |
| Theme | -Current implementation | -
|---|---|
| Dataset slice | One public Xperience-10M sample episode, 5,821 frames, 1,161 windows, and an 8,546-dimensional representation. |
| Modalities | Video, audio, depth, camera pose/SLAM, hand/body mocap, IMU, calibration, and language annotations. |
| Task suite | 20 human-readable tasks form one embodied-AI public-sample suite with shared windowing, split discipline, leakage controls, and minimal/neural head pattern. |
| Baselines | Minimal linear/ridge/logistic heads plus compact PyTorch MLP task heads over the same chronological split; companion simple/NN metadata baselines are also aligned to the selected 128-episode 96/16/16 split. |
| Research directions | Task mapping and extension probes for human modeling, 3D/4D reconstruction, egocentric interaction, and world modeling. |
| Scale-up path | -
-
|
-
| Published mirrors | GitHub repo, GitHub Pages dashboard, GHCR static-site package, HF Space, HF artifact dataset, HF baseline-model repo, and HF collection. |
| Layer | -Current scope | -Where to start | -
|---|---|---|
| Data understanding | -One public Xperience-10M sample episode is converted into 5,821 frames, 1,161 aligned windows, and an 8,546-dimensional multimodal representation. | -PROJECT_BRIEF.md PROJECT_STATUS.md |
-
| Task suite | -
- Twenty human-readable tasks cover recognition, prediction, retrieval, reconstruction, synchronization, long-horizon forecasting, interaction text, action-object binding, sensor bridging, camera sync, and transition timing.
- Historical tier2_task_suite artifact paths are kept for link stability, but they are provenance paths inside the same suite.
- |
-
- 20-task suite note - task contract data - research takeaways - summary report - historical provenance baselines - |
-
| Baselines | -- Minimal heads and compact PyTorch MLP heads provide a controlled single-episode comparison on the same chronological split. - The selected 128-episode setup adds same-split metadata simple/NN baselines for JSON-supported tasks and raw-feature simple/NN baselines on all 20 task axes. - Tasks 15 and 19 are explicitly marked as compact-proxy completions. - | -
- neural MLP outputs - baseline alignment report - raw20 run summary - |
-
| Diagnostics | -Audio contribution, modality ablations, timeline overlays, object labels, and alignment stress tests show which signals are useful and which tasks remain hard. | -audio ablation summary single-episode explorer |
-
| Scale-up | -
-
|
-
- research roadmap - foundation-model plan - selected-128 feature index note - selected-128 feature-index data - selected-128 enhancement note - selected-128 enhancement data - model comparison data - verified Omni result data - Qwen v5/v6 comparison data - Qwen v5/v6 comparison note - Omni model comparison note - verified public package - selected-128 enhancement run - |
-
| Area | -Current decision | -
|---|---|
| Public-sample pipeline | Verified on one public sample episode: 5,821 frames, 1,161 windows, 8,546 dimensions. |
| 20-task suite | Verified minimal baselines with committed metrics, predictions, and manifests. |
| Neural heads | Verified compact PyTorch MLP heads over the same task contracts and chronological splits. |
| Dataset context | Official Xperience-10M links, sample-vs-gated-data boundary, modality coverage, and redistribution policy are documented. |
| Evaluation protocol | Verified generated protocol for windowing, split policy, leakage controls, and per-task metrics. |
| Website and Hub pages | Public dashboard, Hugging Face Space, artifact dataset, baseline model repo, and collection use the same project framing and links. |
| Qwen3-Omni multi-episode pilot | Final verified diagnostic result package exists for the selected 96/16/16 episode split; JSON validity meets the target, while action/subtask metrics remain weak. |
| Raw data / full Qwen weights | Raw Xperience-10M data and full Qwen weights are not redistributed. |
| Step | -Question | -Primary artifacts | -What should be true | -
|---|---|---|---|
| 1 | What is this project? | Project brief Project status Dashboard | A public-sample Xperience-10M research project with 20 tasks, baselines, and a scale-up plan. |
| 2 | What data is used? | Dataset-card alignment Official HF dataset Sample HF dataset | The implemented suite uses one public sample episode; the gated dataset is reserved for selected multi-episode training. |
| 3 | What does one model input contain? | window table feature manifest available-modality data | Each window is an aligned multimodal unit with video, audio, depth, pose/SLAM, mocap, IMU, calibration, and language-derived signals. |
| 4 | What are the 20 tasks? | 20-task suite note task contract data task walkthroughs walkthrough data | Every task has a human-readable name, input, output, metric, baseline scores, and an explicit artifact path. |
| 5 | How are tasks evaluated? | evaluation protocol note evaluation-protocol data | The window unit, chronological split, leakage controls, task metrics, and current limitations are explicit. |
| 6 | What do current results mean? | research takeaways takeaway data summary metrics | Current metrics describe sample-level task behavior and identify which signals need larger held-out experiments. |
| 7 | Which models are implemented? | summary report neural MLP outputs HF baseline repo | Each task has minimal and neural-head evidence over the same feature windows. |
| 8 | What research directions does this support? | research roadmap direction data extension-probe data task contract data | The unified tasks are mapped to human modeling, 3D/4D reconstruction, egocentric interaction, and world modeling. |
| 9 | Which foundation model comes next? | foundation-model plan foundation-model data Native pretraining plan | Qwen3-Omni is the first held-out LoRA baseline; Cosmos 3 has Nano compatibility and Super forward-dynamics LoRA; policy models wait for robot-compatible action targets. |
| 10 | How can the 128-episode suite be pushed without more data? | selected-128 enhancement note selected-128 enhancement data | The enhancement pack proposes dense windows, hierarchical action/subtask labels, raw-feature shard priorities, and multiscale_20s10_40s20_80s40 as the next export target. |
| 11 | How do I reproduce it? | reproducibility guide reproduction audit | Public commands and expected outputs are documented for the sample-episode task suite. |
| 12 | What is still pending? | verified Omni result data data access status multi-episode access status | The final held-out diagnostic Qwen pass is verified and JSON-validity target is met; strong action/subtask model quality remains pending. |
| View | -What to inspect | -Why it matters | -
|---|---|---|
| Project status | project status project-status data | Gives a one-table current project summary before reading the full artifact trail. |
| Data contract | window table feature manifest modality manifests | Confirms what each sample window contains before modeling. |
| Dataset context | dataset-card alignment official dataset links | Explains the official dataset, public sample, modalities, access boundary, and what this repo uses. |
| Visual assets | figure index site assets | Shows the task-suite graphic, modality thumbnails, pipeline diagrams, charts, and logo assets. |
| Evaluation protocol | evaluation protocol note evaluation-protocol data | Defines the task unit, split, metrics, leakage controls, and current limitations. |
| Research roadmap | research roadmap roadmap data | Shows the path from sample-level task development to multi-episode work, larger model tracks, and the future native-pretraining goal. |
| Additional development directions | additional development directions direction data | Records concrete non-backbone tracks: taxonomy, benchmark protocol, representation learning, skill graphs, affordances, 3D/4D memory, QA, and policy transfer. |
| Xperience Embodied Foundation Model plan | native pretraining plan | Describes the long-term full-corpus pretraining goal, target modules, objectives, staged scale-up, hardware ranges, and evaluation protocol. |
| Minimal heads | softmax ridge projection/regression multi-label logistic heads | Keeps every input/output contract visible and inspectable. |
| Neural heads | PyTorch MLP classifiers/regressors under neural_mlp/ | Checks whether nonlinear heads improve each task without changing features. |
| Evidence | metrics predictions confusion matrices diagrams dashboard | Makes the single-episode task development inspectable without rerunning first. |
| Artifact guide | artifact guide | Groups the public evidence after the project overview. |
| Reproducibility contract | reproducibility guide reproducibility matrix | States public commands, expected outputs, exact-match reproduction evidence, and non-reproducible boundaries. |
| Citation metadata | citation metadata software metadata license | Makes the repo easier to cite, index, and reuse without confusing code license and dataset terms. |
