Hi @youbin2014 🤗
I'm Niels and work as part of the open-source team at Hugging Face. I discovered your work through Papers with Code as your paper, SoK: Systematizing LLM Prompt Security: Taxonomies, Datasets, and Unified Evaluation of Attacks and Defenses, recently got featured: https://paperswithcode.co/paper/sok-systematizing-llm-prompt-security.
First of all, congratulations on this amazing systematization work! It's super exciting to see that you have already hosted JailbreakDB on Hugging Face (https://huggingface.co/datasets/youbin2014/JailbreakDB).
I noticed in your GitHub repository (datasec-lab/PromptSecurity) that the companion evaluation-result dataset PromptSecurity-Eval (with the 166,200 harmful-query output records and 9,200 benign-utility records) is currently hosted inside the repository under the dataset/promptsecurity_eval/ folder.
Would you like to also host PromptSecurity-Eval as a dataset on the Hugging Face Hub (https://huggingface.co/datasets)?
Hosting it on Hugging Face would give the evaluation records much more visibility, better discoverability, and would allow researchers to easily explore and load the evaluation results with a single line of code:
from datasets import load_dataset
dataset = load_dataset("datasec-lab/promptsecurity-eval")If you're interested, you can find a quick guide on how to upload datasets here: https://huggingface.co/docs/datasets/loading. Additionally, Hugging Face offers a built-in dataset viewer that lets anyone preview and filter the rows directly from their web browser.
Once uploaded, we can also link both JailbreakDB and PromptSecurity-Eval directly to your Hugging Face paper page (https://huggingface.co/papers/2510.15476) so readers can find all related artifacts in one place.
Let me know if you're interested or if you need any guidance!
Kind regards,
Niels
Hi @youbin2014 🤗
I'm Niels and work as part of the open-source team at Hugging Face. I discovered your work through Papers with Code as your paper, SoK: Systematizing LLM Prompt Security: Taxonomies, Datasets, and Unified Evaluation of Attacks and Defenses, recently got featured: https://paperswithcode.co/paper/sok-systematizing-llm-prompt-security.
First of all, congratulations on this amazing systematization work! It's super exciting to see that you have already hosted JailbreakDB on Hugging Face (https://huggingface.co/datasets/youbin2014/JailbreakDB).
I noticed in your GitHub repository (datasec-lab/PromptSecurity) that the companion evaluation-result dataset PromptSecurity-Eval (with the 166,200 harmful-query output records and 9,200 benign-utility records) is currently hosted inside the repository under the
dataset/promptsecurity_eval/folder.Would you like to also host PromptSecurity-Eval as a dataset on the Hugging Face Hub (https://huggingface.co/datasets)?
Hosting it on Hugging Face would give the evaluation records much more visibility, better discoverability, and would allow researchers to easily explore and load the evaluation results with a single line of code:
If you're interested, you can find a quick guide on how to upload datasets here: https://huggingface.co/docs/datasets/loading. Additionally, Hugging Face offers a built-in dataset viewer that lets anyone preview and filter the rows directly from their web browser.
Once uploaded, we can also link both JailbreakDB and PromptSecurity-Eval directly to your Hugging Face paper page (https://huggingface.co/papers/2510.15476) so readers can find all related artifacts in one place.
Let me know if you're interested or if you need any guidance!
Kind regards,
Niels