Key Questions for CS 376

Neural Computation

Today’s ML systems are a mashup of two kinds of computational objects: the traditional sequential programming that we’re used to is still usually the “outer loop” of an ML system, but that code is the caretaker for a very different kind of animal: a highly parallel vector computer controlled by billions of parameters. This pillar is about understanding how that parallel vector computer works and how we can control it.

ML Systems

Learning Machines

Context and Implications

Notebooks Index
How to Demonstrate Objectives