介绍 CodeAssist，您的个人编码助手，它会随着您的工作而学习。 每次编辑都成为训练数据。每次会话都让它变得更好。 您编码越多，它就越了解您。

The coverage principle: How pre-training enables post-training New preprint where we look at the mechanisms through which next-token prediction produces models that succeed at downstream tasks. The answer involves a metric we call the "coverage profile", not cross-entropy.