Training large language models on narrow tasks can lead to broad misalignment (nature.com) 5 points by thebeardisred 8mo ago ↗ HN
[–] merksittich 8mo ago ↗ Discussions on the manuscript/preprint:https://news.ycombinator.com/item?id=43176553https://news.ycombinator.com/item?id=44554865
1 comment
[ 3.7 ms ] story [ 16.7 ms ] threadhttps://news.ycombinator.com/item?id=43176553
https://news.ycombinator.com/item?id=44554865