SOURCE-LINKED INTELLIGENCE

What FID Hides: Detecting, Ranking, and Diagnosing Deviations in Generative Evaluation

arXiv · AI, language, vision and robotics · article · Aug 25, 2026 · UTC

Generative models are commonly ranked by Fréchet Inception Distance (FID) and Kernel Inception Distance (KID), yet FID's first-two-moment summary can miss distributional differences, and a reported scalar gap alone is not a calibrated test against sampling variation. FID's moment restriction has concrete consequences: on ImageNet, visually unrecognizable images optimized only to match the reference Inception mean and covariance obtain FID $24.7$ versus $58.6$ for held-out real images (lower is better). Moreover, FID and KID are scalar discrepancies that are unchanged when the two samples are e

Read original source ↗ Open in workspace

recordType

paper

region

Global