Ml stuff

What a latent representation is and how it changes throughout model layers What each of the hyperparameters in ultralytics work

L2 Norms, LP Norms, Frobenius Norms, Loss Functions

L2 Weight Decay, dropout, epoch count, learning rate

Connection between matrix rank and how much the model has "learned"

learning rate scheduling

IID sampling from datasets and when to design your dataset to deviate from IID sampling

model bias and variance

boostrapping

bagging