More likely you would just train for emitting svg for some description of a scen... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		svara 8 days ago \| parent \| context \| favorite \| on: Gemini 3 Deep Think More likely you would just train for emitting svg for some description of a scene and create training data from raster images.

		help

recursive 8 days ago [–]

None of this works if the testers are collaborating with the trainers. The tests ostensibly need to be arms-length from the training. If the trainers ever start over-fitting to the test, the tester would come up with some new test secretly.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact