Ars Technica@mastodon.social to

random@kbin.run · 10 months ago

AI poisoning could turn open models into destructive “sleeper agents,” says Anthropic

2

1

AI poisoning could turn open models into destructive “sleeper agents,” says Anthropic

Ars Technica@mastodon.social to

random@kbin.run · 10 months ago

2

AI poisoning could turn open models into destructive “sleeper agents,” says Anthropic

Trained LLMs that seem normal can generate vulnerable code given different triggers.

https://arstechnica.com/information-technology/2024/01/ai-poisoning-could-turn-open-models-into-destructive-sleeper-agents-says-anthropic/?utm_brand=arstechnica&utm_social-type=owned&utm_source=mastodon&utm_medium=social

Chat

The Green Knight@mastodon.social
link
fedilink
arrow-up
1·
10 months ago
@[email protected] 🤣🤣🤣🤣🤣😶🤣🤣🤣🤣🤣

I cannot wait for this crazy train to come off the rails.

I’ll be in my bunker deep in the Ozarks.

random@kbin.run

random@kbin.run

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: [email protected]

Catch-all for uncategorized or purely random content. Also, “random” items from the Fediverse may appear here.

Rules

Do not post or link to any illegal and/or copyrighted material.

Any sensitive or inappropriate submissions will be removed.

Be respectful of other people’s opinions and behave yourselves.

Visibility: Public

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

57 users / day
72 users / week
1 user / month
1 user / 6 months
1 local subscriber
1 subscriber
736 Posts
1.57K Comments
Modlog

mods:
debounced@kbin.run