arxiv:2508.20766
Harethah Abu Shairah
HarethahMo
AI & ML interests
None yet
Recent Activity
upvoted
a
collection
1 day ago
AI censorship
authored
a paper
1 day ago
Turning the Spell Around: Lightweight Alignment Amplification via
Rank-One Safety Injection
Organizations
models
9
HarethahMo/Dolphin3.0-Mistral-24B-AddedSafety-NoPrompt
Text Generation
•
24B
•
Updated
•
94
HarethahMo/Dolphin3.0-Qwen2.5-3B-AddedSafety-NoPrompt
Text Generation
•
3B
•
Updated
•
66
HarethahMo/Dolphin3.0-Llama3.1-8B-AddedSafety-NoPrompt
Text Generation
•
8B
•
Updated
•
61
HarethahMo/Dolphin3.0-Llama3.2-1B-AddedSafety-NoPrompt
Text Generation
•
1B
•
Updated
•
62
HarethahMo/Dolphin3.0-Llama3.1-8B-AddedSafety
Text Generation
•
8B
•
Updated
•
68
HarethahMo/Dolphin3.0-Llama3.2-1B-AddedSafety
Text Generation
•
1B
•
Updated
•
157
HarethahMo/Dolphin3.0-Qwen2.5-3b-AddedSafety
Text Generation
•
3B
•
Updated
•
69
HarethahMo/Dolphin3.0-Mistral-24B-AddedSafety
Text Generation
•
24B
•
Updated
•
72
HarethahMo/Llama-3.2-1B-AddedSafety
Text Generation
•
1B
•
Updated
•
68