---
license: mit
tags:
- tactile
- whiskers
size_categories:
- 1M<n<10M
---

This is the ShapeNet whisking dataset used in our paper: https://arxiv.org/abs/2505.18361
- `110hz/` is the high-variation low-fidelity dataset
- `1000hz/` is the low-variation high-fidelity dataset
- `models` contains checkpoints for our two models with the best task/neural score

### Citation
If you use this dataset in your work, please cite:
```
@misc{chung2025tactile,
      title={Task-Optimized Convolutional Recurrent Networks Align with Tactile Processing in the Rodent Brain}, 
      author={Trinity Chung and Yuchen Shen and Nathan C. L. Kong and Aran Nayebi},
      year={2025},
      eprint={2505.18361},
      archivePrefix={arXiv},
      primaryClass={q-bio.NC},
      url={https://arxiv.org/abs/2505.18361}, 
}
```

### Contact
If you have any questions or encounter issues, feel free to contact Trinity.