I think the most wild thing I've learned about academia is how closed-off everything is. A paper reports that their simulation, that they made themselves, had a 10x speedup. But I can't see their code. How do I know they didn't just mess up the sim?
-
Show this thread
-
Replying to @AhmedMSamara @generativist
Many times I've tried implementing algorithms from papers only to discover huge chunks of logic missing in paper that only become obvious once you try to implement. Of course, no real code provided, which makes one wonder if they ever actually implemented it themselves.
1 reply 0 retweets 3 likes
Replying to @laurence @AhmedMSamara
I’ve had a few simulations I wanted to review with unshared code but were JVm based so I could decompile and yikes were they misspecified.
2:21 PM - 23 Jan 2020
0 replies
0 retweets
1 like
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.