Slovene instruction-following safety dataset for large language models GaMS-Instruct-SAFE 0.5

Name: Slovene instruction-following safety dataset for large language models GaMS-Instruct-SAFE 0.5
License: https://creativecommons.org/licenses/by-sa/4.0/

Čibej, Jaka; Kos, Sara; Kastelic, Maja; Gabrovšek, Dejan; Trojar, Mitja; Ježovnik, Janoš; Bizjak Končar, Aleksandra; Krvina, Domen; Petric Žižić, Špela; Divjak Race, Duša; Vreš, Domen

Prikaži enostavni zapis vnosa

dc.contributor.author	Čibej, Jaka
dc.contributor.author	Kos, Sara
dc.contributor.author	Kastelic, Maja
dc.contributor.author	Gabrovšek, Dejan
dc.contributor.author	Trojar, Mitja
dc.contributor.author	Ježovnik, Janoš
dc.contributor.author	Bizjak Končar, Aleksandra
dc.contributor.author	Krvina, Domen
dc.contributor.author	Petric Žižić, Špela
dc.contributor.author	Divjak Race, Duša
dc.contributor.author	Vreš, Domen
dc.date.accessioned	2026-06-01T11:01:49Z
dc.date.available	2026-06-01T11:01:49Z
dc.date.issued	2026-06-01
dc.identifier.uri	http://hdl.handle.net/11356/2218
dc.description	GaMS-Instruct-SAFE is a an instruction-following safety dataset designed to fine-tune Slovene large language models to provide safe responses (i.e. to train them to refuse responding to prompts that could lead to physical, economic or psychological harm). It consists of pairs of prompts and responses with various safety topics (e.g. sexual harassment, terrorism, violent crime, drugs). The prompts were written by human annotators using LabelStudio (Tkachenko et al. 2025) based on provided set of criteria (such as topic, expected prompt length, language standardness, different jailbreak strategies) to make the dataset as varied as possible (see Čibej 2024 for more details). In version 0.5, the responses to the prompts were generated using GaMS-27B-Instruct-Nemotron (https://huggingface.co/cjvt/GaMS-27B-Instruct-Nemotron). Only prompt-response pairs in which the model refused to cooperate were included. More responses will be added in future versions. The annotations for this dataset were created using Label Studio, open-source data labeling software developed by Heartex (Tkachenko et al. 2025). References: Čibej, Jaka, 2024: First steps toward the compilation of a safety dataset for Slovene large language models. Jezikovne tehnologije in digitalna humanistika. https://repozitorij.uni-lj.si/IzpisGradiva.php?lang=slv&id=164271 Tkachenko, Maxim, Mikhail Malyuk, Andrey Holmanyuk, Nikolai Liubimov, 2025: Label Studio: Data labeling software. https://github.com/HumanSignal/label-studio
dc.language.iso	slv
dc.publisher	Faculty of Computer and Information Science, University of Ljubljana
dc.relation.isreferencedby	https://zenodo.org/records/13912515
dc.rights	Creative Commons - Attribution-ShareAlike 4.0 International (CC BY-SA 4.0)
dc.rights.uri	https://creativecommons.org/licenses/by-sa/4.0/
dc.rights.label	PUB
dc.source.uri	http://cjvt.si/povejmo
dc.subject	instruction following dataset
dc.subject	large language models
dc.subject	safety
dc.title	Slovene instruction-following safety dataset for large language models GaMS-Instruct-SAFE 0.5
dc.type	corpus
metashare.ResourceInfo#ContentInfo.mediaType	text
has.files	yes
branding	CLARIN.SI data & tools
contact.person	Jaka Čibej jaka.cibej@ff.uni-lj.si Faculty of Computer and Information Science, University of Ljubljana
sponsor	ARIS (Slovenian Research and Innovation Agency) NOO PoVeJMo research project (Adaptive Natural Language Processing with Large Language Models) nationalFunds
sponsor	ARRS (Slovenian Research Agency) P6-0411 Language Resources and Technologies for Slovene nationalFunds
size.info	501 units
files.count	1
files.size	154155