Alejandro Álvarez Espejo

Limitations of Conducting Symbolic Reasoning Research via Chatbot InterfacesLack of Model ControlResearchers cannot fix model parameters (e.g., temperature, sampling strategy), which introduces ran...

Lab Note No. 1: Why We're Testing AI with ContradictionsMost AI research focuses on performance—fluency, usefulness, accuracy. This project starts elsewhere: not in what AI can do, but in how it be...