Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
jyhong
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
1.
▲
by
jyhong
11mo ago
Paper link: https://arxiv.org/abs/2510.13928 TL;DR: We show that continued training on junk social posts makes LLMs think worse. This big concern here is that because many teams keep topping up models with web crawls.
2.
▲
by
jyhong
2y ago
It is not labor-intensive if you want a prompt to work in 80%-90% of cases and humans are good at that. But it is labor-intensive if you want to make it to work at 99%. Then you need to go through many cases and "optimize" the pro