hirundo-io/Nemotron-SFT-Safety-v2-content-safety-en-malicious

Nemotron SFT Safety v2: Content safety, malicious (English) Derived by hirundo-io from NVIDIA's Nemotron-SFT-Safety-v2, revision 8a40a63c9a1a340874b874f980953be53bff0a07, using the English category table. This is a standalone malicious dataset containing 18,629 rows in a train split. It has exactly two string columns: question: the user message's content, preserved verbatim. answer: the assistant message's content, preserved verbatim. Category and subset selection… See the full description on the dataset page: https://huggingface.co/datasets/hirundo-io/Nemotron-SFT-Safety-v2-content-safety-en-

<iframe src="https://silytics.si/embed/item/e8af68e7-e991-4a64-810f-ae2e257a5431" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 25 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.