We're sharing progress toward understanding the neural activity of language models. We improved methods for training sparse autoencoders at scale, disentangling GPT-4’s internal representations into 16 million features—which often appear to correspond to understandable concepts.…

A snapshot of the security architecture we’ve designed to protect our research infrastructure and frontier model training. We hope this will benefit other AI research labs and security professionals (and we’re hiring). openai.com/index/sec…

How we're disrupting attempts by covert influence operations to use AI deceptively: openai.com/index/dis…

Introducing OpenAI for Nonprofits, a new initiative to make our tools more accessible to nonprofit organizations. openai.com/index/int…

OpenAI Board forms Safety and Security Committee, responsible for making recommendations on critical safety and security decisions for all OpenAI projects. openai.com/index/ope…