the fact that we need a separate tool to audit healthcare ml models for safety and trust is just... https://www.reddit.com/user/hypergraphr