From hard refusals to safe-completions: toward output-centric safety training - OpenAI

hard.indah.link

[unable to retrieve full-text content]

From hard refusals to safe-completions: toward output-centric safety training  OpenAI The Link Lonk


August 08, 2025 at 12:03AM
https://news.google.com/rss/articles/CBMiXEFVX3lxTE9UbW9KdFZ0dnJsOE5NVFlMRTFzREFHMnpwUDdoSW45X0tIYXFlTGN2NFJTQl9LTVdISE9FSEdJemQ1aW1wUjUtbDhQMnB2a0NZQzVpelprLWNld24z?oc=5

From hard refusals to safe-completions: toward output-centric safety training - OpenAI

https://news.google.com/search?q=hard&hl=en-US&gl=US&ceid=US:en