A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models
Paper • 2605.08513 • Published • 16
An Charles Yang
charlesy1
AI & ML interests
None yet
Recent Activity
upvoted a paper 11 days ago
A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models upvoted a paper 11 days ago
π-Bench: Evaluating Proactive Personal Assistant Agents in Long-Horizon WorkflowsOrganizations
None yet
upvoted 3
papers 11 days ago