Artificial Fintelligence

January 2025

October 2024

Papers I've read this week: vision language models

They kept releasing VLMs, so I kept writing...

Oct 28, 2024 • 

Finbarr Timbers

37

July 2024

Papers I've read this week

This is a grab bag of papers.

Jul 10, 2024 • 

Finbarr Timbers

11

4

March 2024

How does batching work on modern GPUs?

The first and most important optimization you can do for any modern deep learning system, generally speaking, is to implement batching.

Mar 1, 2024 • 

Finbarr Timbers

32

3

January 2024

Where do LLMs spend their FLOPS?

LLM theory, with a hint of empirical work

Jan 29, 2024 • 

Finbarr Timbers

30

1

December 2023

The evolution of the LLM API market

Before I studied machine learning, I was an Econ grad student banging out OLS problem sets (I see the OLS equation— (X’X)^-1X’y— whenever I close my…

Dec 13, 2023 • 

Finbarr Timbers

28

13

The evolution of the LLM API market

Note: if you’re coming to this post online, this is the same as the free post, I ran into issues opening this article up on Substack.

Dec 12, 2023 • 

Finbarr Timbers

6

November 2023

Transformer inference tricks

How to make your model run faster than a greased pig

Nov 23, 2023 • 

Finbarr Timbers

32

7

October 2023

Why do LLMs use greedy sampling?

"Greedy sampling is the worst form of sampling, except all those other forms that have been tried from time to time." - Winston Churchill, if he worked…

Oct 17, 2023 • 

Finbarr Timbers

13

4

September 2023

More on Mixture of Experts models

6 papers on different routing mechanisms

Sep 7, 2023 • 

Finbarr Timbers

35

1