Viewing profile — jyhong

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 ke…

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 ne…