They are trained on these "benchmarks", that's why they score better. | Hacker News

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

		coliveira 1 day ago \| parent \| context \| favorite \| on: Huawei releases an open weight model trained on Hu... They are trained on these "benchmarks", that's why they score better.

elzbardico 1 day ago [–]

If they were trained on those benchmarks they would score 100%

coliveira 1 day ago | [–]

They show how bad they are, they cannot score 100% on benchmarks they were trained on.

Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4
Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact