Adversarial Pretraining Trade-offs in Deep Tabular Models: Robustness vs. In-Distribution Performance

This report synthesises findings from 16 peer-reviewed papers addressing the following research question: Does the incorporation of synthetic adversarial examples during pretraining lead to trade-offs between in-distribution performance and robustness on TabRobust, and how does this compare to. 8 claims were extracted from source literature; 8 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 8.3/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research goal: Does the incorporation of synthetic adversarial examples during pretraining lead to trade-offs between in-distribution performance and robustness on TabRobust, and how does this compare to traditional data augmentation techniques?

Autonomous literature synthesis. Automated review score: 8.3/10. Full text and citation available at Assignee Research.