The interactive film is in the book's Companion

Play it, pause it, turn it and drag it with Mathematics for AI and Machine Learning.

Get the book

Mixture of Experts: Routing Tokens to a Few Blocks

Chapter 10 · Mathematics for AI and Machine Learning · Advanced · 1:56

An interactive live film: four score planes over the token plane, their upper envelope picks each token's experts (top-k with softmax gates). Drag the router directions until every expert carries its share of the load.

Narration

A mixture of experts layer has several expert blocks, here four, and a router in front of them. The dots on the floor are tokens; to draw them, they live in two dimensions.

The router gives every expert a score, a linear function of the token: h i equals w i dot x, where w i is a row of the router matrix. Above the floor, each score is a tilted plane, and the arrow on the floor is its direction w i.

The rest of the narration is in the book's Companion.

More from Chapter 10

Interactive2:11Tensor Parallelism: Block Products Across DevicesChapter 10Preview1:24Block Diagonal Matrices: Separate WorldsChapter 10Preview1:28Block Multiplication: Blocks Multiply Like NumbersChapter 10Preview1:39Longer Contexts with Rotary PositionsChapter 10Preview2:20Woodbury Identity: a Rank-One Update of the InverseChapter 10Preview1:40Schur Complement: What Is Left After Minimizing Out a BlockChapter 10Preview1:32Kronecker Product and the Vectorization IdentityChapter 10Preview1:40Block Cholesky: Two Steps to a CircleChapter 10Preview1:30Online Softmax, One Tile at a TimeChapter 10