ShrinkageLens: Information Sources in Perturbation Discrimination Scores

Perturbation discrimination ranks held-out responses against a prediction, often after magnitude correction. A corrected score can, however, obtain perturbation-specific information from scoring inputs instead of the prediction. ShrinkageLens separates these sources by specifying each rule's inputs and comparing scores using identical predictions, matched permutations and the inverse of a known shrinkage map. Identical predictions score exactly $(N+1)/(2N)$ among $N$ candidates under official ties for nine rules without row-specific inputs. Scaling each prediction against its own truth raises the shared response from near 0.51 to 0.80-0.84, whereas own-truth gene weights lower it. On five Perturb-seq panels, magnitude-stratified permutations raise most null means but move no own-truth ${\ell }_1$ score across its range. Of 78 published model-context pairs, 32 beat the mean and 28 of these exceed their global matched null (25 under strata). GEARS on one RPE1 split and scBERT, whose predictions are effectively constant, do not. Inverting an unclipped empirical-Bayes map on positive-weight coordinates restores the unshrunk ${\ell }_2$ score within $10^{-3}$ in all 90 settings. Scalar recovery varies with panel, center and weights. These controls locate a score's information source without ranking corrections by biological validity.