An Overview of Local Explainability Methods and Their Evaluation in AI

An Overview of Local Explainability Methods and Their Evaluation in AI

UofU Data Science via YouTube Direct link

Simulatability

12 of 15

12 of 15

Simulatability

Class Central Classrooms beta

YouTube videos curated by Class Central.

Classroom Contents

An Overview of Local Explainability Methods and Their Evaluation in AI

Automatically move to the next video in the Classroom when playback concludes

  1. 1 Announcements
  2. 2 It's hard to opt-out
  3. 3 Confidence scores
  4. 4 Explanations in plain English free-text / chain-of-thoughts
  5. 5 Input attribution gradient-based & select-then-predict
  6. 6 Feature interactions effective attention
  7. 7 Concept-based explanations TCAV
  8. 8 Data influence influence functions
  9. 9 Contrastive explanations contrastive editing
  10. 10 Explainability as a dialog
  11. 11 Taxonomy of evaluation of explanations
  12. 12 Simulatability
  13. 13 Why are application-grounded evals of explanations scarce in NLP?
  14. 14 Application-grounded evaluations
  15. 15 Trust in AI

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.