Follow the Gradient: Tracing Safety Degradation Back to Training Data
Paper • 2610.37481 • Published • 30
Ari Morcos
arimorcos
AI & ML interests
None yet
Recent Activity
upvoted a paper 1 day ago
Follow the Gradient: Tracing Safety Degradation Back to Training Data upvoted a paper about 1 year ago
BeyondWeb: Lessons from Scaling Synthetic Data for Trillion-scale
PretrainingOrganizations
upvoted a paper 1 day ago