Hi @xiangan 🤗
Niels here from 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/2605.25979.
The paper page lets people discuss about your paper and lets them find artifacts about it (your models, datasets or demo for instance), you can also claim
the paper as yours which will show up on your public profile at HF, add Github and project page URLs.
It's fantastic to see that you've already released several models and datasets on the 🤗 Hub! I noticed that these artifacts are not yet linked directly to your paper page (https://huggingface.co/papers/2605.25979). Linking them will greatly improve their discoverability and visibility, allowing researchers and practitioners to easily find and cite your work.
Here are the models and datasets that are already on the Hub but could be linked:
Models:
Datasets:
I also saw that you plan to release the LLaVA-OneVision-2-4B-Instruct model and introduced the new JumpScore temporal-localization benchmark dataset. It would be awesome to have these also available on the 🤗 Hub once they are ready, so that people can do:
from datasets import load_dataset
dataset = load_dataset("your-hf-org-or-username/your-dataset")We can help you link the existing models and datasets to the paper page (read here).
For the pending LLaVA-OneVision-2-4B-Instruct model and JumpScore dataset, we can provide guidance on uploading them. See here for a models guide: https://huggingface.co/docs/hub/models-uploading, and for datasets: https://huggingface.co/docs/datasets/loading.
Let me know if you're interested/need any guidance!
Cheers,
Niels
ML Engineer @ HF 🤗
Hi @xiangan 🤗
Niels here from 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/2605.25979.
The paper page lets people discuss about your paper and lets them find artifacts about it (your models, datasets or demo for instance), you can also claim
the paper as yours which will show up on your public profile at HF, add Github and project page URLs.
It's fantastic to see that you've already released several models and datasets on the 🤗 Hub! I noticed that these artifacts are not yet linked directly to your paper page (https://huggingface.co/papers/2605.25979). Linking them will greatly improve their discoverability and visibility, allowing researchers and practitioners to easily find and cite your work.
Here are the models and datasets that are already on the Hub but could be linked:
Models:
LLaVA-OneVision-2-8B-Instruct: https://huggingface.co/lmms-lab-encoder/LLaVA-OneVision-2-8B-InstructOneVision-Encoder: https://huggingface.co/lmms-lab-encoder/onevision-encoder-largeOneVision-Encoder-Lang: https://huggingface.co/lmms-lab-encoder/onevision-encoder-large-langDatasets:
LLaVA-OneVision-2-VideoCaption: https://huggingface.co/datasets/mvp-lab/LLaVA-OneVision-2-Data/tree/main/mid_training_videoLLaVA-OneVision-2-Spatial: https://huggingface.co/datasets/mvp-lab/LLaVA-OneVision-2-Data/tree/main/spatialI also saw that you plan to release the
LLaVA-OneVision-2-4B-Instructmodel and introduced the newJumpScoretemporal-localization benchmark dataset. It would be awesome to have these also available on the 🤗 Hub once they are ready, so that people can do:We can help you link the existing models and datasets to the paper page (read here).
For the pending
LLaVA-OneVision-2-4B-Instructmodel andJumpScoredataset, we can provide guidance on uploading them. See here for a models guide: https://huggingface.co/docs/hub/models-uploading, and for datasets: https://huggingface.co/docs/datasets/loading.Let me know if you're interested/need any guidance!
Cheers,
Niels
ML Engineer @ HF 🤗