A key insight: central flows can predict loss graphs across various neural network architectures. They also offer a more satisfying explanation for why complex optimizers like RMSProp are effective. 📈 #OptimizerTheory 3/6
PAC-Bayes tells you where the cat’s been.
Kakeya tells you where it can go.
We use cone crossings and directional bounds to track optimizer paths. Tighter than PAC-Bayes. No flat priors.
Wanna know how?
New ADAM paper drops soon.
#CatsOfML #Kakeya #OptimizerTheory #CSTheory
