Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: High-Res Neural Cellular Automata

cells2pixels.github.io

Post reply on HN
Print view View original

51–56 of 56 posts

Re: Show HN: High-Res Neural Cellular Automata

#51

Post by bjourne »

yorwba wrote: post #49
bjourne wrote: post #46

Earlier quoted context omitted.

Nope. Many comments from green accounts get flagged and then auto-killed unless vouched for.

This is a green account that got auto-killed: https://news.ycombinator.com/item?id=48304852 This is a green account where someone used the "flag" button to kill it: https://news.ycombinator.com/item?id=48262928 Note how the former only says [dead] while the latter is [flagged] [dead].

The spam filter does both. You can verify for yourself that flagging a comment most of the time is not enough to kill it.

Re: Show HN: High-Res Neural Cellular Automata

#52

Post by yorwba »

bjourne wrote: post #51
yorwba wrote: post #49

Earlier quoted context omitted.

This is a green account that got auto-killed: https://news.ycombinator.com/item?id=48304852 This is a green account where someone used the "flag" button to kill it: https://news.ycombinator.com/item?id=48262928 Note how the former only says [dead] while the latter is [flagged] [dead].

The spam filter does both. You can verify for yourself that flagging a comment most of the time is not enough to kill it.

The spam filter influences the number of times a comment needs to be flagged before it becomes [flagged] and gets killed, but if it says [flagged], someone used the "flag" button at least once. Sometimes just downvoting a spam comment is enough to kill it; those just get marked [dead] and not [flagged].

Re: Show HN: High-Res Neural Cellular Automata

#55

Post by Akimsa3 »

That's really impressive i've been thinking that we have using technologies cause they were a breakthrough but we actual resources we can implement technollogy that was relegated time ago, like when implementing 3d there were many other until the one who was more consistent is how we visualize 3d today. Cause this looks pretty simillar to metaballs. An idea that is not longer that hard pursued but i just saw the similarity.

Re: Show HN: High-Res Neural Cellular Automata

#56

Post by WhiteNoiz3 »

esychology wrote: post #38
WhiteNoiz3 wrote: post #29

Earlier quoted context omitted.

IIRC training starts with the initial state and the end state, and the end state is always oriented the same way. It would be interesting to see what would happen if the end state was rotated randomly though I suspect it wouldn't work so well.

Checkout the Isotropic NCA blog from the Google Zurich team: https://google-research.github.io/self-organising-systems/is...

interesting, thanks!
Post reply on HN