A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models
Paper • 2605.08513 • Published • 15
Atoosa Chegini
AtoosaChegini
AI & ML interests
None yet
Recent Activity
authored a paper 15 days ago
A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models upvoted a paper 15 days ago
A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language ModelsOrganizations
None yet
authored a
paper 15 days ago
authored 5
papers about 2 years ago
Run-Off Election: Improved Provable Defense against Data Poisoning Attacks
Paper • 2302.02300 • Published
Robustness of AI-Image Detectors: Fundamental Limits and Practical Attacks
Paper • 2310.00076 • Published • 2
EditVal: Benchmarking Diffusion Based Text-Guided Image Editing Methods
Paper • 2310.02426 • Published
Fast Adversarial Attacks on Language Models In One GPU Minute
Paper • 2402.15570 • Published • 2
What do we learn from inverting CLIP models?
Paper • 2403.02580 • Published • 4