OpenAI is set to release its first open-weight language model with reasoning capabilities since GPT-2 in the coming months, according to CEO Sam Altman.
An open-weight language model has publicly accessible trained parameters, allowing developers to analyze and fine-tune the model for specific tasks without needing the original training data. These models differ from open-source models, which provide complete access to the source code, training data, and methodologies.
Altman mentioned that the company will engage with developers to discuss how to maximize the utility of the open-weight language model. The first event is scheduled to take place in San Francisco within a few weeks, followed by sessions in Europe and the Asia-Pacific region.
"We still have some decisions to make, so we are hosting developer events to gather feedback and later test early prototypes," he noted.
In February, Altman outlined a new roadmap for the latest models. Microsoft-backed OpenAI aims to transition to a for-profit company by the end of the year to secure the full $40 billion funding led by SoftBank Group.
The startup has indicated that this transition is necessary to obtain the capital required for developing advanced AI models.