Are the weights invertible, or are the prompts being fed into the model invertib... | Hacker News

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

		kobelb 3 months ago \| parent \| context \| favorite \| on: Language models are injective and hence invertible Are the weights invertible, or are the prompts being fed into the model invertible?

singularity2001 3 months ago [–]

The prompt is reconstructable from the (latent) representation vector but obviously not from the output

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