Explainability, Reasoning, Priors and GPT-3
Machine Learning Street Talk (MLST)
00:00
Understanding Model Explainability
This chapter explores the challenges of machine learning model explainability, specifically focusing on decision trees and their limitations. It highlights the complexities of relying on simplistic metrics and emphasizes the risks of interpreting decisions based on arbitrary classifications. The discussion also critiques the potential for misleading simplicity in explanations and advocates for more fluid approaches in model design.
Transcript
Play full episode
