gpt-oss-safeguard technical report

OpenAI has developed two reasoning models, gpt-oss-safeguard-120b and gpt-oss-safeguard-20b, which were post-trained from the gpt-oss models and designed to reason according to a provided policy. This report describes the capabilities of gpt-oss-safeguard and provides baseline safety evaluations.

Summary written by Kernelia from the original article by OpenAI News. The story and its rights belong to its author.