Musings by Chris Hayduk
Subscribe
Sign in
Home
The Sciences
The Humanities
The Rest
Archive
About
Latest
Top
Discussions
The Foundation Model Trap
Why AI Model Companies Are More Like Airlines than Like Cereal Companies
Mar 5
•
Chris Hayduk
5
Share this post
Musings by Chris Hayduk
The Foundation Model Trap
Copy link
Facebook
Email
Notes
More
2
Understanding DeepSeek Part II: DeepSeek-V2
Compressing the key-value matrix
Mar 5
•
Chris Hayduk
3
Share this post
Musings by Chris Hayduk
Understanding DeepSeek Part II: DeepSeek-V2
Copy link
Facebook
Email
Notes
More
February 2025
A Primer on Multi-Head Causal Self-Attention
The neural network layer that kicked off the LLM craze
Feb 1
•
Chris Hayduk
1
Share this post
Musings by Chris Hayduk
A Primer on Multi-Head Causal Self-Attention
Copy link
Facebook
Email
Notes
More
2
January 2025
Understanding DeepSeek Part I: DeepSeekMoE
Mixture of experts models with a twist
Jan 30
•
Chris Hayduk
4
Share this post
Musings by Chris Hayduk
Understanding DeepSeek Part I: DeepSeekMoE
Copy link
Facebook
Email
Notes
More
Understanding Protein Language Models Part III: Structure Prediction without Multiple Sequence Alignment in ESMFold
How ESMFold and ESM3 replace explicit MSAs with encoder-only transformers
Jan 22
•
Chris Hayduk
3
Share this post
Musings by Chris Hayduk
Understanding Protein Language Models Part III: Structure Prediction without Multiple Sequence Alignment in ESMFold
Copy link
Facebook
Email
Notes
More
Understanding Protein Language Models Part II: Encoder-only Transformers as Continuous Fuzzy String Matching
How transformers learn from their input data
Jan 15
•
Chris Hayduk
3
Share this post
Musings by Chris Hayduk
Understanding Protein Language Models Part II: Encoder-only Transformers as Continuous Fuzzy String Matching
Copy link
Facebook
Email
Notes
More
December 2024
OpenAI o3 and the Rise of the Intelligence Allocator
The implications of rapidly increasing inference costs
Dec 20, 2024
•
Chris Hayduk
1
Share this post
Musings by Chris Hayduk
OpenAI o3 and the Rise of the Intelligence Allocator
Copy link
Facebook
Email
Notes
More
On Algorithmic Moats and the Path to AGI
Google's path to winning the AI race
Dec 19, 2024
•
Chris Hayduk
1
Share this post
Musings by Chris Hayduk
On Algorithmic Moats and the Path to AGI
Copy link
Facebook
Email
Notes
More
July 2024
Rejuvenating the Political System
How do we get the US government unstuck?
Jul 22, 2024
•
Chris Hayduk
Share this post
Musings by Chris Hayduk
Rejuvenating the Political System
Copy link
Facebook
Email
Notes
More
June 2024
ESM3 and the Future of Protein Language Models
Pure sequence learning is out, multiscale data is in
Jun 25, 2024
•
Chris Hayduk
3
Share this post
Musings by Chris Hayduk
ESM3 and the Future of Protein Language Models
Copy link
Facebook
Email
Notes
More
1
A Perspective on the Limitations of Language Modeling
Probing the upper limits of compute required for AGI
Jun 22, 2024
•
Chris Hayduk
Share this post
Musings by Chris Hayduk
A Perspective on the Limitations of Language Modeling
Copy link
Facebook
Email
Notes
More
4
A Case Study in Finetuning Open Source LLMs: Training LLaMA 2 for the Text-to-SQL Task
Introduction
Jun 4, 2024
•
Chris Hayduk
Share this post
Musings by Chris Hayduk
A Case Study in Finetuning Open Source LLMs: Training LLaMA 2 for the Text-to-SQL Task
Copy link
Facebook
Email
Notes
More
Share
Copy link
Facebook
Email
Notes
More
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts