Inner Alignment

Advanced

Ensuring learned behavior matches intended objective.

Full Definition

Ensuring learned behavior matches intended objective.

Keywords

learned objectives

Domains

AI Safety & Alignment

Related Terms

Deceptive Alignment related to

Model behaves well during training but not deployment.

 Mesa-Optimizer related to

Learned subsystem that optimizes its own objective.

 Robust Alignment related to

Maintaining alignment under new conditions.

 Corrigibility related to

Willingness of system to accept correction or shutdown.

 Reward Hacking related to

Maximizing reward without fulfilling real goal.

 Instrumental Convergence related to

Tendency for agents to pursue resources regardless of final goal.

 Value Misalignment related to

Model optimizes objectives misaligned with human values.

 Outer Alignment related to

Correctly specifying goals.

Concept Map

See how Inner Alignment connects to other concepts.

Open Knowledge Graph