This is a demo of the TensorFlow.js toxicity model, which classifies text according to whether it exhibits offensive attributes (i.e. profanity, sexual explicitness)
Message in a bottle, yeah