“central claim is that for larger LLMs, most neurons represents a singular concept and things clustered around that concept, perhaps with small offshoots and some use of low-level activations.”

Posted on by

https://thezvi.substack.com/p/i-can-see-everything

Beyond predicting the next token. I remember smart people being very confident about what an LLM was. The experts always seemed less confident.

“Take a neuron that tracks ‘a sense of righteousness when a bargain purchase turns out to work better than the expensive alternative’. Neurons that track a feeling or social cue like this are some of the most common we’ve found”