A Guide To Linearity and Nonlinearity in Machine Learning

-

…and their roles in decision boundaries, embeddings, dynamical systems, and next-gen LLMs

“A watch for an eye fixed, a tooth for a tooth.”
Lex Talionis, Codex Hammurabi

The famed Lex Taleonis is a law of proportionality. You’re taking my eye, I take yours. You’re taking my tooth, I take yours (being a Babylonian dentist will need to have been tough).

The law was not put in place to foster violence; quite, it aimed to limit it. The Lex Taleonis envisioned a legal world where every thing might be described by linear equations: every crime would create an output proportional to its input. And because the punishment for an offense was proportional to the crime, it avoided excessive retribution and explosions of violence that left every thing of their wake destroyed.

Beyond the world of retribution, linearity plays a very important role in our serious about the world: in linear systems, every thing is known. There isn’t any chaos, no complicated maths. All scientists would need to do all day was solve these sorts of equations:

For each motion, there’s an equal and opposite response.
Newton’s Third Law of Motion

Unfortunately, the fact we inhabit is way from this linear utopia. History is ripe with examples of the world responding to small things in highly disproportional ways: the Defenestration of Prague sparking the Thirty…

ASK ANA

What are your thoughts on this topic?
Let us know in the comments below.

0 0 votes
Article Rating
guest
0 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments

Share this article

Recent posts

0
Would love your thoughts, please comment.x
()
x