Improved methods for training sparse autoencoders at scale. Shows promise towards interpreting GPT-4's neural activity: https://t.co/zNbYnYr6lE

— Greg Brockman (@gdb) 6 juin 2024

Improved methods for training sparse autoencoders at scale. Shows promise towards interpreting GPT-4's neural activity:

→ View original post on X — @gdb