Machine Learning

Artificial Intelligence

Neural Networks

Interpretability

Explore recommended reads, summarized by AI

Understanding AI Systems

Explorations of how machine learning models work internally and methods for explaining their decisions.

Large Language Model Internals

Model Explainability Tools

Popular this week

Gabrieljearle

Interpretability Through The Lens of Fractal Plants | Gabriel Earle

Sajjad Pakdamansavoji

OpenAI’s gpt-6-astra Probed Network Weights Without Tools. The Logs Say Less Than the Headline

In

Towards AI

by

Vedant Pandhare

Mechanistic Interpretability: We Built the Most Powerful Minds in History. We Can't Read Them.

Kevin O'Shaughnessy

Claude’s take on Neural Networks Are Already Clean

Shmulik Cohen

Evil Qwen Has Jewish Cousins

Latest stories

kaleem

Weight Space Programming Explained: An LLM Read a Network’s Weights, Then Rewrote Them

S. Salmon

We broke a language model on purpose. Could Claude find the damage and fix it by hand?

Sajjad Pakdamansavoji

OpenAI’s gpt-6-astra Probed Network Weights Without Tools. The Logs Say Less Than the Headline

Uday Kumar Boya

Nobody Fully Knows How AI Thinks. Here's the Field Trying to Find Out

Ashley

One AI Fixed Another by Reading Its Mind

See more stories

Who to follow

Popular this week

Gabrieljearle

Interpretability Through The Lens of Fractal Plants | Gabriel Earle

Sajjad Pakdamansavoji

OpenAI’s gpt-6-astra Probed Network Weights Without Tools. The Logs Say Less Than the Headline

In

Towards AI

by

Vedant Pandhare

Mechanistic Interpretability: We Built the Most Powerful Minds in History. We Can't Read Them.

Kevin O'Shaughnessy

Claude’s take on Neural Networks Are Already Clean

Shmulik Cohen

Evil Qwen Has Jewish Cousins

Latest stories

kaleem

Weight Space Programming Explained: An LLM Read a Network’s Weights, Then Rewrote Them

S. Salmon

We broke a language model on purpose. Could Claude find the damage and fix it by hand?

Sajjad Pakdamansavoji

OpenAI’s gpt-6-astra Probed Network Weights Without Tools. The Logs Say Less Than the Headline

Uday Kumar Boya

Nobody Fully Knows How AI Thinks. Here's the Field Trying to Find Out

Ashley

One AI Fixed Another by Reading Its Mind

See more stories

Who to follow