Explainer

Banks were scoring credit risk with models in the 1980s. A history of machine learning in finance, and why explainability came first.

Google researchers asked 53 practitioners about their data. Ninety-two percent had watched a data problem surface much later in a live system.

Credit scoring is named as high-risk in the AI Act itself. What the law demands of a lender, and which rulebooks apply at the same time.

Self-attention weighs every token at once, a convolution slides across pixels. Why one system writes a summary and another spots a tumor.

Correct the labels in ImageNet and ResNet-18 overtakes ResNet-50. Benchmarks are treated as ground truth and almost never audited.