Training Compute-Optimal Large Language Models

训练计算最优的大型语言模型.

Locating and Editing Factual Associations in GPT

定位和编辑GPT中的事实关联.

Modifying Memories in Transformer Models

修正Transformer模型中的记忆.

Towards TracIng Factual Knowledge in Language Models Back to the Training Data

将语言模型中的事实知识追溯到训练数据.

On the Role of Bidirectionality in Language Model Pre-Training

探讨语言模型预训练中的双向性.

使用pydensecrf构造条件随机场

Building fully-connected conditional random field with pydensecrf.