Model inspection and interpretation at Seldon
Practical AI
00:00
Unlocking Model Interpretability with Alibi
This chapter explores Alibi's innovative black box explanation methods that enhance model interpretability, regardless of architecture. It discusses key techniques like 'anchor' and LIME, emphasizing their role in providing clear insights into machine learning predictions and the vision for Alibi as a comprehensive explanation library.
Play episode from 27:59
Transcript
