Mnestic probing recovers signal where amnesic removal found none

measured in 1 paper

Rozanova et al. introduce mnestic probing: project onto the union of INLP-trained classifiers' rowspaces (keep only the probe directions), the mirror image of amnesic removal, using the same trained classifiers [rozanova-etal-2023-interventional] Fed incrementally into the NLI head, it produces a clear monotonic accuracy increase for the composite and gold-label features, faster than a random-direction control [rozanova-etal-2023-interventional] It succeeds exactly where the amnesic (removal) version found no effect, because it operates in a low-rank regime that sidesteps the low-class-count control problem [rozanova-etal-2023-interventional] Context monotonicity's increase is not clearly above the random baseline, so the paper does not claim it is used [rozanova-etal-2023-interventional]

Structure

Linear Subspace

Context

mnestic probing (union-of-rowspaces projection), complement of INLP nullspace projection

Confirmed in models

RoBERTa-large-MNLI · RoBERTa, RoBERTa-large-MNLI + HELP fine-tuning · RoBERTa, RoBERTa-large-MNLI + double fine-tuning (HELP + NLI_XY) · RoBERTa, BERT-base-uncased fine-tuned on SNLI · BERT, BERT-base-uncased fine-tuned on SNLI + HELP · BERT

Method

INLP subspace + AlterRep counterfactual push

Papers

Interventional Probing in High Dimensions: An NLI Case Study — Rozanova, Julia, Valentino, Marco, Cordeiro, Lucas, Freitas, André2023