PolicyIK

Paper | Code | Project Page

PolicyIK is a goal-conditioned reinforcement learning policy for high-precision inverse kinematics. This repository holds its simulation policies for the 7-DoF OpenArm. In simulation, a single attempt keeps the mean errors over the final second within 1 mm and 0.2° on 93.5% of held-out tasks, and a deployment protocol raises success to 98.9%. The training data and evaluation tasks are in PolicyIK-Data.

Contents

Path Description
checkpoints/policyik_s{1,2,3}.pth Policies from training seeds 1, 2 and 3. The files are identical to simulation/checkpoints/ in the code repository.

Usage

from huggingface_hub import hf_hub_download

path = hf_hub_download("tina390/PolicyIK", "checkpoints/policyik_s1.pth", revision="v1.0")

Then evaluate it with the code repository:

cd simulation
bash scripts/evaluate.sh 1 /path/to/policyik_s1.pth

License

Apache-2.0, the same as the code repository.

Citation

@article{liu2026policyik,
  title   = {PolicyIK: High-Precision Inverse Kinematics for Robot Arms with Reinforcement Learning},
  author  = {Liu, Yuhan and Liu, Xinqi and Dong, Zixuan and Ross, Keith W. and Wan, Shenji},
  journal = {arXiv preprint arXiv:2611.05372},
  year    = {2026}
}
Downloads last month

-

Downloads are not tracked for this model. How to track
Video Preview
loading

Dataset used to train tina390/PolicyIK

Paper for tina390/PolicyIK