A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models
Paper • 2605.08513 • Published • 17
🔄 In a Training Loop
Manzi Fabrice Niyigaba
fabricemanzi
·
AI & ML interests
NLP
Recent Activity
upvoted a paper 20 days ago
A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language ModelsOrganizations
upvoted a paper 20 days ago