N0-VTLA - NeoSim Pour Ball
Task policy for 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.
| 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 pour_ball task.
100%, first attempt, at simulator commit 695a22d. Under the current branch tip's tightened
criterion the same checkpoint scores 65% - the tightened version additionally requires the ball
to have left the cup. Both numbers are real; they answer different questions.
Serving
VTLA_ASSET_ID=univtac_pourball_joint_norm \
python scripts/serve_zmq.py --config sim_single_arm_tactile \
--ckpt <this-dir> --addr "tcp://127.0.0.1:5557"
action_horizon is 50; set exec_horizon: 50. Prompt: pour ball
Evaluation protocol
Measured on the UniVTAC simulator at commit 695a22d
(branch NeoSim of 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 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.