Minimax-optimal Scaling Laws in Ridge Regression via Transfer Learning
Transfer from ridge-regression pretraining can achieve minimax-optimal scaling under power-law covariance shift.
2 stories, newest first · one card per story
Transfer from ridge-regression pretraining can achieve minimax-optimal scaling under power-law covariance shift.
Theory shows a fixed recurrent memory matching current predictions can still be insufficient for future learning.
Today 1.8K readers · 4K views · 3K reads · all time 162K views