petemetefete

↗ HN profile [ 19.9 ms ] full profile

Karma: 0
Created: ()
Submissions: 0

Training large language models on narrow tasks can lead to broad misalignment (nature.com)

3 points 6mo ago 1