Deliberative alignment: reasoning enables safer language models

Company news OpenAI Source language: English

Deliberative alignment: reasoning enables safer language models Introducing our new alignment strategy for o1 models, which are directly taught safety specifications and how to reason over them.

Excerpt

Deliberative alignment: reasoning enables safer language models Introducing our new alignment strategy for o1 models, which are directly taught safety specifications and how to reason over them.

Short excerpt from the source — open the original for the full text.

Sources

← Back to the daily brief