Completed
The explanation is always true; It is what the model thinks.
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
How to Fail Interpretability Research
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Premeditation of evils
- 3 Interpretability hype.
- 4 well... we've been here before...
- 5 What this talk is about
- 6 What this talk is NOT about
- 7 We need interpretability to increase user trust.
- 8 We need to understand every single bit of the model.
- 9 Agenda Many opportunities to fail.
- 10 We first must define a universal mathematical definition of interpretability
- 11 The performance and interpretability trade-off is inevitable.
- 12 How I present the explanations doesn't
- 13 Since there is no good way to evaluate interpretability methods, I can only show you qualitative results
- 14 I am a computer scientist! Running human
- 15 The explanation is always true; It is what the model thinks.
- 16 I'm just a researcher who provide technical tools. The real world usage is something I cannot control.
