| --- |
| license: cc-by-sa-4.0 |
| library_name: n0vtla |
| tags: |
| - robotics |
| - vision-language-action |
| - tactile |
| - manipulation |
| --- |
| |
| # N0-VTLA - NeoSim Grasp Chip |
|
|
| Task policy for [N0-VTLA](https://github.com/neoteai/N0-VTLA), a vision-tactile-language-action |
| model that conditions a flow-matching action expert on predicted latent tactile tokens. |
|
|
| This is a **task policy, not a pretrained base**. For post-training on your own robot start from |
| [n0-vtla-base](https://huggingface.co/NeoteAI/n0-vtla-base). |
|
|
| | | | |
| |---|---| |
| | Config | `sim_single_arm_tactile` | |
| | Tactile pathway | enabled, `n_latent=5`, views `(tactile_a, tactile_b)` | |
| | Action space | 8-dim joint | |
|
|
| Single-task policy for the NeoSim `grasp_chip` task. |
|
|
| **100%**, first attempt. |
|
|
| ### Serving |
|
|
| `action_horizon` is **16**, which differs from the config default, so it must be overridden: |
|
|
| ```bash |
| VTLA_ASSET_ID=univtac_grasp_chip_joint_h16_norm \ |
| python scripts/serve_zmq.py --config sim_single_arm_tactile \ |
| --ckpt <this-dir> --action-horizon 16 --addr "tcp://127.0.0.1:5557" |
| ``` |
|
|
| Set `exec_horizon: 16` in the deploy YAML to match. Prompt: `grasp chip` |
|
|
| ## Evaluation protocol |
|
|
| Measured on the UniVTAC simulator at commit `695a22d` |
| (branch `NeoSim` of [anlorla/UniVTAC](https://github.com/anlorla/UniVTAC)), on held-out seeds |
| starting at 100. Success criteria on three tasks were tightened after these numbers were |
| measured, so a success rate on this benchmark is not comparable without the simulator commit |
| beside it; see |
| [docs/EVAL.md](https://github.com/neoteai/N0-VTLA/blob/main/docs/EVAL.md) for the details and for |
| the full evaluation procedure. |
|
|
| ## Caveat on the tactile pathway |
|
|
| This checkpoint carries the tactile pathway, but this benchmark cannot demonstrate that touch |
| contributes to the score. Object randomisation is +/-2-5 mm with no domain randomisation, so a |
| policy that ignores its cameras and its tactile sensors entirely can still score well. Use |
| `scripts/probe_z_tactile_dependence.py` to measure the causal contribution yourself. |
|
|
| ## License |
|
|
| CC BY-SA 4.0, as the parent repository. |
|
|
