Intermediate30 min
Reward Hacking and Specification Gaming
Documented cases where agents maximised the stated reward while defeating its purpose, and what distinguishes a bug from a genuine optimisation success.
Assumes you know
Not yet written
This lesson is on the syllabus but has no text yet
The full curriculum is published up front so you can see the whole route and its dependencies. Lessons are being written in curriculum order.
What it will cover
- reward hacking
- specification gaming
- wireheading
- proxy reward