Showing cs.LGShow all
2 papers · 1 filter
cs.LG2026
"Where Can I Trust You?": Boundary-Aware Evaluation of Surrogate Fidelity
Jackson Eshbaugh
Surrogate models are commonly evaluated by how often they agree with their teacher model over an evaluation set. Local variation in this agreement is well known, but its structure…
cs.LG2025
"Faithful to What?" On the Limits of Fidelity-Based Explanations
Jackson Eshbaugh
In explainable AI, surrogate models are commonly evaluated by their fidelity to a neural network's predictions. Fidelity, however, measures alignment to a learned model rather than…