Enhancing Genomic Foundation Model Robustness through Iterative Black-Box Adversarial Training
TL;DR * We test a genomic foundation model (DNABERT-2 encoder with a lightweight classifier head) using a black-box genetic algorithm that proposes biologically plausible edits constrained by GC balance, motif preservation, and transition bias. * With only 2 to 3 single-nucleotide substitutions in 300 base pairs, the attack succeeds on...
Oct 14, 20258