Conceptobservational attribution
Observational Attribution
An attribution approach measured using correlation coefficients between feature activations and outputs, linear probe coefficients, or rank statistics; it requires no modification of internal variables and can only capture naturally co-varying relationships.
Timeline (last 90 days)
Oct 14
For a large class of response functions, observational feature discovery evidence generally cannot identify intervention attribution, which is a structural conclusion rather than a flaw of any specific method.
Unverified50%