gpt-oss-safeguard-120b and gpt-oss-safeguard-20b are two open-weight reasoning fashions post-trained from the gpt-oss fashions and skilled to purpose from a offered coverage with a view to label content material below that coverage. On this report, we describe gpt-oss-safeguard’s capabilities and supply our baseline security evaluations on the gpt-oss-safeguard fashions, utilizing the underlying gpt-oss fashions as a baseline. For extra details about the event and structure of the underlying gpt-oss fashions, see the unique gpt-oss mannequin mannequin card.
Source link
Article Categories:
Water Purifiers & Accessories