Measuring Goodhart’s law

New GPT-3 capabilities: Edit & insert

A research agenda for assessing the economic impacts of code generation models

Economic impacts research at OpenAI

Lessons learned on language model safety and misuse

Solving (some) formal math olympiad problems

Aligning language models to follow instructions

Introducing text and code embeddings

Text and code embeddings by contrastive pre-training

WebGPT: Improving the factual accuracy of language models through web browsing

Customizing GPT-3 for your application

OpenAI Residency