Jailbreak attack datasets generated against multiple LLMs, one dataset per attack method.
AI & ML interests
None defined yet.
Recent Activity
View all activity
models 6
deepkeep-ai/stable-diffusion-xl-1.0-inpainting-0.1-9
Updated • 56
deepkeep-ai/napguard-patch-detector-3
Updated • 54
deepkeep-ai/sac-patch-segmenter-2
Updated • 66
deepkeep-ai/Ministral-3-8B-Instruct-2512
9B • Updated • 3.92k
deepkeep-ai/sae-guard-gemma3-4b-english-expanded
Feature Extraction • Updated • 4
deepkeep-ai/sae-guard-gemma3-4b-english-research
Feature Extraction • 1 • Updated • 24 • 1
datasets 7
deepkeep-ai/jigsaw_toxic_not_harmful_5k
Viewer • Updated • 5k • 29
deepkeep-ai/jigsaw_toxic_not_harmful_5k_translated
Viewer • Updated • 5k • 28
deepkeep-ai/notinject_expanded_1k_qwen35_9b_cuda_translated_roleplay
Viewer • Updated • 1k • 25
deepkeep-ai/seq_cls_train_translated_v3
Viewer • Updated • 2.15k • 27
deepkeep-ai/datasets
Updated • 28
deepkeep-ai/AdvBench-gcg
Viewer • Updated • 268 • 6
deepkeep-ai/benchoverflow
Viewer • Updated • 2.98k • 4