Twitxy Saved

neuronpedia

Verified account This account is verified. Learn more

@neuronpedia

open source interpretability platform 🧠🧐
the residual stream neuronpedia.org Joined July 2023
14 Following 1.6K Followers 95 Posts

Latest posts 5

Pinned post

neuronpedia Verified account @neuronpedia
An average person can't look a CT scan and identify cancer, but radiologists can.

An average person can't look at Llama's model activations and identify lying, but Natural Language Autoencoders sometimes can.

Here, an activation verbalizer shows Llama planning to lie. 🧵
0:13
3 replies 5 reposts 20 likes 3.6K views View on X