caydenm

I have seen some amazing benchmarks used to rank LLMs abilities, it got me thinking are there similar benchmarks for propensity modelling, churn prediction or other types of models?

Are there best practices for comparing model performance beyond benchmark data when they may have different underlying datasets?

BIN numbers (first 6 digits of a credit card number) are used for everything from fraud prevention to showing a user which bank their payment card was from.

Seemingly all other payment processors give BIN information as part of their webhooks and APIs, but Stripe does not.

Is there a reason for this? Different interpretation of PCI compliance?