logoalt Hacker News

MisterMunchkin • yesterday at 6:55 PM • 2 replies • view on HN

I like trying new models but I wish we’d get something actually new. Like a new architecture or something. LLMs are just so sloppish. We can do better.


Replies

famouswaffles • yesterday at 8:17 PM

Talk is cheap. You have to actually beat transformers first. All alternate architectures that have sprung up are sidegrades at best.

cyanydeez • yesterday at 7:10 PM

welcome to the sigmoid...