heystack

Specification gaming examples in AI - master list

Gaming examples of the AI finding unintended solutions to the specified objective that don't satisfy the designer's intent. This includes reward hacking behaviors in reinforcement learning.

From the heystacks.com archiveUpdated 2.5K views
Open original

Change Requests

No proposals on this page. Improvements start with a suggested change.

Discussion 0

0/5,000

No discussion yet. Start it — no sign-up needed.