Release Terminal Bench Pro on Hugging Face

Hi @yangcole 🤗

I'm Niels and work as part of the open-source team at Hugging Face. I discovered your work through Hugging Face's daily papers as yours got featured: https://huggingface.co/papers/2512.24873.
The paper page lets people discuss about your paper and lets them find artifacts about it, and you can also claim the paper as yours which will show up on your public profile.

It's great to see that you've already hosted the **ROME** model on the Hub (`FutureLivingLab/iFlow-ROME`)!

Would you also like to host the **Terminal Bench Pro** dataset and the training trajectories you've released on https://huggingface.co/datasets? Hosting them on Hugging Face will give them more visibility and better discoverability within the agentic AI community. It also allows researchers to easily load the data using the `datasets` library:

```python
from datasets import load_dataset

dataset = load_dataset("FutureLivingLab/terminal-bench-pro")
```

Besides that, there's the [dataset viewer](https://huggingface.co/docs/hub/en/datasets-viewer) which allows people to quickly explore the benchmark tasks and trajectories directly in the browser.

After uploaded, we can also link the dataset to the paper page (read [here](https://huggingface.co/docs/hub/en/model-cards#linking-a-paper)) so people can discover all your work in one place.

Let me know if you're interested or need any guidance!

Kind regards,

Niels

Review the Terminal Bench Pro dataset and training trajectories referenced in the issue, then read the linked Hugging Face dataset viewer and paper-linking documentation. Done means the dataset and trajectories are hosted on Hugging Face, loadable with the datasets library, viewable in the browser, and linked from the paper page.

Written by the indexing model from the issue text.