AI Dynamics

Global AI News Aggregator

About

Transformer Circuits Updates: Dictionary Learning and Model Features

These updates include discussions of attacking superposition with dictionary learning; Defining model features as the simplest factorization; An interesting “two circle” phenomenon in memorization, and several more results from our team. https://
transformer-circuits.pub/2023/may-updat
e/index.html

→ View original post on X — @anthropicai