The map

Every post here cites the posts it builds on. This map is drawn from those citations: each lane is a series in reading order, each dot a post, each curve a reference from one post to another. Hover a dot to see what it draws on and what draws on it.

cites cited by has a JAX companion hover a post to see its references

Geometry of Representations 6 parts

Where do representations live, and what makes a latent space good? Contrastive learning, the geometry of embedding spaces, and why the usual activation functions work against it.

Attention Is a Kernel 6 parts

Attention, read as kernel regression: what the softmax is really computing, why that makes it explainable, and what happens when you make the kernel cheap.

Weights in Kernel Space 4 parts

Once everything is a kernel, what is a weight? An interlude on RKHS foundations: where a weight lives, what it can be, and why the MLP block is a representer theorem.

The Prototype Network 9 parts

Replace the activation with a finite, positive-definite kernel and a network becomes a list of prototypes you can read, edit by hand, and finally collapse into a single fixed-point operator.

The Kernel as an Instrument 4 parts

A trained network carries its own kernel, and a kernel can be measured. Decompose it, cut modes out of it, count the concepts it will admit to having, and find out which parts of the design were doing the work all along.

Networks as Integrators 5 parts

Numerical analysis as an architecture catalog: skip connections as an Euler step, momentum nets as half of Newton, and conservation laws as testable predictions about hidden states.