arXiv:2606. 21806v2 Announce Type: replace Abstract: Deep generative models reproduce the observational distribution of their training data, inheriting any spurious associations it contains.
By Jingyuan Chen, Kangrui Ruan, Junzhe Zhang
arXiv:2607. 23388v1 Announce Type: cross Abstract: As constrained learning becomes increasingly common, models are trained under explicit feasibility requirements to enforce fairness, safety, robustness, regulariza- tion, and physics or logic constraints.
By Xin Wang (Jeff), R. Tyrrell Rockafellar (Jeff), Xuegang (Jeff), Ban
arXiv:2609.14124v1 Announce Type: cross
Abstract: Medical image analysis is often hindered by biased datasets, which can lead to biased models and limited clinical applicability. A promising strategy...
By Yasin Ibrahim, Robin J. Evans, Konstantinos Kamnitsas
The paper introduces the Directional Influence Function (DIF), a new estimator for measuring how training samples affect the solution of constrained learning problems. DIF incorporates constraints directly by formulating optimality conditions as a variational inequality, thereby avoiding the bias seen in classical influence functions. Experiments on constrained linear regression and fairness‑constrained CNNs show that DIF accurately predicts changes in test loss and aligns closely with actual retraining results.
By Xin Wang (Jeff), R. Tyrrell Rockafellar (Jeff), Xuegang (Jeff), Ban
arXiv:2604. 11704v2 Announce Type: replace-cross Abstract: Deep Neural Networks are highly susceptible to shortcut learning, frequently memorizing low-dimensional spurious correlations instead of underlying causal mechanisms.
By Nicolas Rodriguez-Alvarez (Instituto de Educacion Secundaria Parquesol, Valladolid, Spain)
The paper introduces CauVaDE, a causal generative model that represents an unobserved confounder as a discrete latent cluster, enabling accurate approximation of both observational and interventional image distributions. It demonstrates that this representation can capture diverse causal explanations without sacrificing fit to the data. A novel causal bootstrap method is proposed to vary confounder entropy, producing a family of distinct interventional image distributions.
By Jingyuan Chen, Kangrui Ruan, Junzhe Zhang