All the materials from our work: From Specialization to Generalization: Instruction-tuned LLMs for Robust Harmful Content Mitigation
Lukas Edman
leukas
AI & ML interests
NLP, tokenization, pretraining, low-resource
Recent Activity
updated a collection 2 days ago
HIPPO updated a collection 2 days ago
HIPPO updated a collection 2 days ago
HIPPOOrganizations
models 174
leukas/hippo-32b
24B • Updated
leukas/hippo-4b
4B • Updated
leukas/Qwenfant-0.4B
0.4B • Updated • 11
leukas/Qwenfant-0.1B
0.1B • Updated • 9
leukas/mbblm_en_only
0.1B • Updated • 3
leukas/amlm_hd_fail
Fill-Mask • 34.7M • Updated • 2.84k • 1
leukas/amlm_hard_nhot
Fill-Mask • 50M • Updated • 5.45k
leukas/amlm_hard_decay
Fill-Mask • 34.7M • Updated • 2.82k
leukas/amlm_hard
Fill-Mask • 34.7M • Updated • 5.4k
leukas/byt5-small-nc16-10k-ruen
0.3B • Updated • 9
datasets 10
leukas/hippo_superset
Viewer • Updated • 738k • 9
leukas/Qwenfood
Viewer • Updated • 272k • 63 • 1
leukas/climbmix-1b-shuffle
Viewer • Updated • 2.06M • 100
leukas/climbmix-4b-shuffle
Viewer • Updated • 8.21M • 214
leukas/cute_zs
Viewer • Updated • 14k • 8
leukas/hatexplain-filtered
Viewer • Updated • 6.92k • 12
leukas/Wiktionary
Viewer • Updated • 301k • 13
leukas/GramGen
Viewer • Updated • 126 • 12
leukas/cute_results
Viewer • Updated • 16 • 4
leukas/cute
Viewer • Updated • 14k • 175 • 2