Update README.md

README.md

@@ -1,3 +1,92 @@
 ---
 license: mit
+tags:
+  - robotics
+  - tactile
+  - visuo-tactile
+  - representation-learning
+library_name: pytorch
 ---
+
+# NeoForce: A Unified Force Tactile Representation Model
+
+[Technical Report](https://research.neoteai.com/assets/n0-foundation-paper.pdf) · [Project Website](http://research.neoteai.com/n0-foundation) · [Code](https://github.com/SparkleXFantasy/N0-Foundation-Preview) · [OpenNeoData](https://huggingface.co/datasets/NeoteAIEmbodied/OpenNeoData)
+
+Tactile hardware is fragmented, so a model built on one signal format is bound to the device that produced it. NeoForce instead describes every tactile observation as a **dense three-axis force field** over the sensing surface, capturing shear and pressure in a form that is physically grounded and shared across sensors, and learns a temporally structured representation on top of it.
+
+The model is a joint visual-tactile ViT: RGB frames and dual-sensor force fields are patched onto one shared token grid and run through a single trunk, so the modalities attend to each other directly instead of being fused after the fact. A temporal transformer over per-frame CLS pairs supplies context, and a progressive decoder reconstructs the force field and contact mask at the full input resolution, supervised at every scale it passes through.
+
+This repository holds the weights. The code, together with the data preparation, training and evaluation guides, lives in the [code repository](https://github.com/SparkleXFantasy/N0-Foundation-Preview).
+
+## Files
+
+| File | Size | What it is |
+|---|---|---|
+| `neoforce/neoforce.pt` | 1.25 GB | the NeoForce model |
+| `visuo_tactile_conversion/visuo_tactile_conversion.pt` | 49 MB | the tactile conversion model |
+
+### `neoforce/neoforce.pt`
+
+The trained NeoForce model at step 100,000, kept as a full training state rather than inference weights alone:
+
+| Key | Contents |
+|---|---|
+| `model` | the student: the visual-tactile encoder plus the JEPA predictor, 156.14 M parameters |
+| `teacher` | the EMA copy the JEPA and DINO targets are drawn from |
+| `model_cfg` | the model configuration, so nothing has to be restated to rebuild it |
+| `data_cfg`, `force_norm` | the per-channel scale the force targets were divided by |
+| `step` | the training step the weights come from |
+
+### `visuo_tactile_conversion/visuo_tactile_conversion.pt`
+
+The frozen tactile conversion model, which turns raw tactile camera frames into the force field NeoForce consumes:
+
+```
+gel image -> bird-view correction -> optical flow based on the reference frame
+          -> conversion network -> masked, scaled force field
+```
+
+## Setup
+
+Each file has a fixed path the code looks in:
+
+```bash
+git clone https://github.com/SparkleXFantasy/N0-Foundation-Preview
+cd N0-Foundation-Preview/neoforce
+
+mkdir -p weights neoforce/visuo_tactile_conversion/model
+curl -L -o weights/neoforce.pt \
+  https://huggingface.co/NeoteAI/NeoForce/resolve/main/neoforce/neoforce.pt
+curl -L -o neoforce/visuo_tactile_conversion/model/visuo_tactile_conversion.pt \
+  https://huggingface.co/NeoteAI/NeoForce/resolve/main/visuo_tactile_conversion/visuo_tactile_conversion.pt
+```
+
+The [Installation Guide](https://github.com/SparkleXFantasy/N0-Foundation-Preview/blob/main/neoforce/docs/Installation.md) covers the rest of the setup.
+
+## Model
+
+| | |
+|---|---|
+| Backbone | ViT-B, 12 layers, 768-dim, patch 20 |
+| Input | 4 frames @ 30 fps, RGB 360×640 + tactile representation `(6, 360, 640)` |
+| Output | tactile representation `(6, 360, 640)`, contact mask `(6, 360, 640)`, global force `(6,)` |
+
+The six force channels are `[left fx, fy, fz, right fx, fy, fz]`, one triplet per tactile sensor.
+
+Trained on [OpenNeoData](https://huggingface.co/datasets/NeoteAIEmbodied/OpenNeoData), the 5,000-hour open-source subset of NeoData.
+
+## Citation
+
+```bibtex
+@misc{n0foundation,
+      title={N0-Foundation: Towards the Age of Tactile Intelligence},
+      author={NeoteAI Team and TEAI Team},
+      year={2026},
+      url={https://research.neoteai.com/assets/n0-foundation-paper.pdf},
+      note={Technical Report}
+}
+```
+
+## License
+
+[MIT](https://opensource.org/licenses/MIT)