Publications

Abstract

Credit-risk models are trained on proxy labels and deployed under temporal and segment change, yet no single transfer metric separates base-rate shift, probability-scale shift, and feature-label relationship change. We contribute a design-science artifact: a locked, multi-signal audit protocol for supervision drift in proxy-labeled credit-risk prediction. Five layers (transfer performance, an oracle-gap probe, a calibration diagnostic, feature-label stability, and a synthetic positive control), thresholds, and decision rules were locked before interpretation; a bounded reading is a designed outcome. On a public LendingClub dataset (temporal 2013 to 2016 and cross-segment transfer), ranking is stable and oracle gaps are small; the clearest temporal signal is a prevalence and probability-scale mismatch that intercept-only diagnostic recalibration largely reduces, though its cause is not identifiable from the available release. The positive control responds only to larger injected shifts; subtler drift cannot be excluded. Mapping diagnostic patterns to governance actions is conceptual guidance, not validated here.

Abstract

AI evaluation can support the wrong inference when an in-domain benchmark success does not survive distribution shift, or when the benchmark endpoint is entangled with a design factor. We study this problem in CRISPRi perturbation-effect prediction, evaluating a frozen Geneformer representation under a locked, pre-registered protocol: heads and model selection were frozen before test evaluation; the protocol required external outcome labels to remain withheld until final unblinding; and analysis-governing decisions were fixed before the evaluations they govern. In-distribution on the Virtual Cell Challenge (VCC), the frozen representation carries measurable predictive information beyond a dimension-matched random-feature control (ΔR² = +0.1645, 95% CI [+0.1375, +0.1920]), satisfying the pre-registered informativeness gate required before interpreting transfer. It then fails zero-shot transfer on both external screens (Spearman ρ = -0.139 and -0.267), lying below that control on each. Adding a predefined magnitude block improves the representation externally (Δρ = +0.032 and +0.143) but, under the frozen primary head, does not rescue transfer: both remain negative. A pre-registered, count-adjusted max-response secondary is positively associated with the outcome on both screens; we report it as correlational and secondary, not as a recovered magnitude signal. Finally, the VCC endpoint is strongly sample-size associated: a count-only linear model reaches R² = +0.4325, versus +0.2589 for the four magnitude scalars; adding those scalars to cell count improves R² by only +0.0017, so much of the aggregate-magnitude signal overlaps with cell count. This case study shows how locking the evaluation, harmonizing the measured endpoint, and separating primary from secondary evidence can change the inference supported by an AI benchmark.

Abstract

Transcriptomic foundation models are increasingly used as reusable cell and gene representations, but validating them on new data under weak supervision and distribution shift is hard: standard comparisons conflate genuine representation signal with model capacity, row-identity artifacts, gains over strong task-specific baselines, and outcome rules chosen after seeing the test set. We introduce a pre-registered, final-test-once evaluation framework that locks the outcome rule, seeds, and target-gene-grouped splits before any test data are seen, and scores each frozen representation against a strong expression baseline, a matched-capacity Gaussian control, and a within-split row-identity (shuffle) control; only the per-cell embedding-extraction step is model-specific. Applying it to three architecturally distinct models—Geneformer, scGPT, and UCE—across two external Replogle Perturb-seq datasets (RPE1 and K562), all three clear the capacity and row-identity controls by a wide margin, yet none reliably beats the expression baseline: the strongest (Geneformer) exceeds it by at most about 0.03 test R² and clears the pre-registered four-of-five-seed threshold in neither dataset, while scGPT and UCE fall below it. All three therefore land in the same pre-registered partial-replication category—a consistent cross-architecture outcome, even though the baseline-relative gap differs in sign and magnitude across models. These representations carry real structure beyond trivial controls but, under this weak magnitude label, do not transfer past a simple strong baseline; the locked framework is reusable for any frozen transcriptomic representation by swapping only the extraction step.

Abstract

Learning from weak, proxy, or relative supervision is common when ground-truth labels are unavailable, but model reliability under distribution shift remains challenging. This paper introduces supervision drift, defined as changes in the relationship between inputs, weak labels, and contexts. Using CRISPR-Cas13d transcriptomic perturbation experiments, we construct a controlled non-IID benchmark across cell lines and timepoints. The results show meaningful in-domain learning and partial cross-cell-line transfer, but major failures under temporal shift, highlighting the need for feature-stability diagnostics before deployment.