Completed
Issue #4: Evaluations of robustness are not robust
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Challenges in Fostering Trust and Distrust in AI Systems
Automatically move to the next video in the Classroom when playback concludes
- 1 Human trust in AI recap
- 2 What factors enable people to trust trustworthy models?
- 3 Challenges of constructing valid benchmarks: Data shortcuts
- 4 Issue #2: Evaluations with independent test instances
- 5 Issue #3: Not ensuring we are not breaking the basics
- 6 Issue #4: Evaluations of robustness are not robust
- 7 Challenges of ensuring high-quality pretraining data
