A research paper details how decomposing groups of neurons in a neural network into interpretable "features" may improve safety by enabling monitoring of LLMs (Anthropic)

Anthropic:
A research paper details how decomposing groups of neurons in a neural network into interpretable “features” may improve safety by enabling monitoring of LLMs  —  Neural networks are trained on data, not programmed to follow rules.  With each step of training …

from Techmeme https://ift.tt/rxnXVSc