How can they be sure that the results are actual words from the scroll and not j... | Hacker News

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

		Fleusal on Oct 13, 2023 \| parent \| context \| favorite \| on: First word discovered in unopened Herculaneum scro... How can they be sure that the results are actual words from the scroll and not just hallucinations of the neural network? If the scrolls are in such bad condition that the data is almost only noise then what stops their high-tech deep learning model from just.. making it all up?

lolc on Oct 14, 2023 [–]

As long as the model doesn't know about words, and finds letter sequences that correspond to words, we can conclude that it found actual writing.

Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10
Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact