Dataset · Hugging Face

ZJU-Safety/DataShield

Name: DataShield
Creator: ZJU-Safety
License: other
Keywords: task_categories:text-generation, language:en, language:es, language:ru, language:de, language:pl, language:th, language:vi, language:sv, language:bn, language:da, language:he, language:it, language:fa, language:sk

🛡️ DataShield: Uncovering Risky Fine-Tuning Data Across LLMs Through Consensus Subspace Alignment ⚡ Find risky data before fine-tuning. ⚡ We use DataShield to continuously release risk-scored versions of widely used fine-tuning datasets. Every release keeps the original training example together with one final risk_score, making it easy to remove the highest-risk portion before training. Quick Start · Choose a Ratio · Code · Paper ✨ Overview Each record… See the full description on the dataset page: https://huggingface.co/datasets/ZJU-Safety/DataShield.

View on Hugging FaceReport an issue

Author / organization	ZJU-Safety
License (as declared)	other
Downloads	160 (as reported by Hugging Face)
Likes	1
Created	26 August 2026
Last updated at source	9 October 2026

safetyalignmentinstruction-tuningdata-filteringrisk-scoringemnlp

Embed this entry

<iframe src="https://silytics.si/embed/item/cf3d7cc9-f9ce-4b45-9f56-d4b0e79b2a15" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 16 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.