Fuck AI@lemmy.world · 5 months ago

A small number of samples can poison LLMs of any size

www.anthropic.com

cross-posted to:
fuck_ai@lemmy.world

A small number of samples can poison LLMs of any size

www.anthropic.com

pyre@lemmy.world to

Fuck AI@lemmy.world · 5 months ago

cross-posted to:
fuck_ai@lemmy.world

Anthropic research on data-poisoning attacks in large language models

Seems like an invitation to me.

Archive link: https://web.archive.org/save/https%3A%2F%2Fwww.anthropic.com%2Fresearch%2Fsmall-samples-poison

Chat

marius@feddit.org
link
fedilink
arrow-up
7·
5 months ago
Doesn’t that mean that the model overfits?

Fuck AI@lemmy.world

fuck_ai@lemmy.world

Create a post

You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !fuck_ai@lemmy.world

“We did it, Patrick! We made a technological breakthrough!”

A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.

AI, in this case, refers to LLMs, GPT technology, and anything listed as “AI” meant to increase market valuations.

Visibility: Public

This community can be federated to other instances and be posted/commented in by their users.

1.48K users / day
4.78K users / week
8.46K users / month
16.8K users / 6 months
2 local subscribers
6.43K subscribers
3.31K Posts
48.2K Comments
Modlog