Cite
Notes
Only stored in your browser.
Attribution
Bilinear MLPs enable weight-based mechanistic interpretability
arXiv 2024
from 1 papers
Alice Rigg
Lee Sharkey
Michael T. Pearce
Thomas Dooms