Residue-Level Attributions in Protein Allergenicity Models Do Not Recover Allergen Epitopes

Deep allergenicity classifiers are increasingly used in safety screening of novel foods, and recent protein language models have substantially improved protein-level allergenicity prediction. However, whether their explanations capture biologically meaningful information remains unclear. We introduce an epitope-grounded residue-level benchmark for quantitatively evaluating attribution faithfulness in protein allergenicity models. Across frozen ESM-2, multi-task ESM-2, and DeepPlantAllergy, protein-level classification was robust, yet classification-head explanation signals did not significantly exceed random in residue-level alignment with annotated epitopes across AUROC, AUPRC, and Precision@$k$. Integrated Gradients identified residues that were functionally important to the model, but did not overlap annotated epitopes. Saturation mutagenesis further suggested that classifiers rely predominantly on physicochemical and compositional sequence features rather than epitope-specific mechanisms. Residue-level importance signals should therefore not be interpreted as immunological explanations for safety screening or hypoallergen design without quantitative validation.