Improving Model Safety Behavior with Rule-Based Rewards

公司动态 OpenAI 原文语言:英文

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

原文摘录

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

以下是来源提供的简短摘录,完整内容请查看原文。

参考来源

← 返回每日简报