You wouldn’t happen to be german, would you?
- 2 Posts
- 8 Comments
- Pudutr0n@lemmy.worldtoAsk Lemmy@lemmy.world•What's your entirely unimportant and inconsequential hot take?0·4 days ago
- Pudutr0n@lemmy.worldtoAsk Lemmy@lemmy.world•What's your entirely unimportant and inconsequential hot take?1·4 days ago
This sounds reasonable
- Pudutr0n@lemmy.worldtoAsk Lemmy@lemmy.world•What's your entirely unimportant and inconsequential hot take?5·4 days ago
walruses should be called saber-toothed seals
please keep the convo on track. we’re talking about burritos
- Pudutr0n@lemmy.worldOPtoShowerthoughts@lemmy.world•Someone should make a community to freely distribute examples of data poisoning people can randomly put in their social media posts/images to sabotage AI1·6 days ago
I’ve read in papers that you can poison datasets with a very small percentage of the data, if done cleverly. I can fish up the source if you want (but it might take me some time).
edit: here it is.
We conduct the largest pretraining poisoning experiments to date, pretraining models from 600M to 13B parameters on chinchilla-optimal datasets (6B to 260B tokens). We find that 250 poisoned documents similarly compromise models across all model and dataset sizes (…)
Emphasis mine. All it takes is 250 poisoned documents.
Oh, THAT guy. I replied to him a few times about a week ago asking him if he was OK, what he was angry about and such and then he DMd me. I didn’t reply to that, and then he got banned.
If you’re weird like me and want to go deeper into the lore of this person, he claimed he was doing this because he started some anti-ai (art?) community and it got a lot of traction, only to get removed within a day by admins. He claimed the people that reported him were from lemmy shitpost (i think… maybe lemmy.world).
Did i miss the latest disgruntled user drama due to real world responsibilities again?
I really should get my priorities straight.
are those terminator teeth?