Agentic Scientific Debugger: Detecting Hidden Failures in ML Experiments

Another project, Another Blog.Recently, I started thinking about a simple question. How do we actually debug machine learning experiments? When we think about debugging, we usually think about fixin