I don't trust papers out of “Top Labs” anymore
old.reddit.com
I don't trust papers out of “Top Labs” anymore
1–10 of 53 posts
Re: I don't trust papers out of “Top Labs” anymore
#2"Jeff Dean spent enough money to feed a family of four for half a decade to get a 0.03% improvement on CIFAR-10."
Nailed it.
Re: I don't trust papers out of “Top Labs” anymore
#3No post body was provided.
Re: I don't trust papers out of “Top Labs” anymore
#4Quoted post unavailable.
I actually find that smaller communities on Reddit are pretty good.
Just unsub from all the defaults and find your place.
Talking about Reddit as a whole is not easy. It’s insane how wide it is.
Re: I don't trust papers out of “Top Labs” anymore
#5Quoted post unavailable.
What would be a good place to post your misgivings about a general research area?
Re: I don't trust papers out of “Top Labs” anymore
#6It does seem like better algorithms to get similar results from smaller models should be prioritised.
Rather than throwing more compute at a problem for 0.03 better score, show me one tenth the compute with a loss of 0.03 score. That would be impressive and far more useful.
Re: I don't trust papers out of “Top Labs” anymore
#7Quoted post unavailable.
Where the discussion is infinitely deeper and more interesting than your throwaway comment here…
Re: I don't trust papers out of “Top Labs” anymore
#8Quoted post unavailable.
No true Scotsman posts anywhere you disagree with!
Your logical fallacy is not the flex you think it is.
Re: I don't trust papers out of “Top Labs” anymore
#9No post body was provided.