Improving Model Safety Behavior with Rule-Based Rewards

Ignore

OpenAI Blog · 2024-07-24 09:00 UTC

Not analyzed yet

Eligible for automatic cleanup in 2 day(s) unless marked Must Read.

Content

We’ve developed and applied a new method leveraging Rule-Based Rewards (RBRs) that aligns models to behave safely without extensive human data collection.

Open original article ↗

Your feedback

Keep this article

Protects it from automatic cleanup.