logoalt Hacker News

jbellis • today at 7:26 PM • 4 replies • view on HN

It tops benchmarks because it uses them in its training data. https://x.com/eliebakouch/status/2077425801633427919

Replies

krackers • today at 8:20 PM

training on the test set is all you need

➕ show 1 reply
sva_ • today at 7:31 PM

https://nitter.net/eliebakouch/status/2077425801633427919

That is very disappointing.

➕ show 1 reply
AtlasBarfed • today at 9:41 PM

scrape the scrapers... all's fair...

sajithdilshan • today at 7:46 PM

Disappointing. Sounds like Volkswagen 2.0

➕ show 1 reply