Futurology Today
  • Communities
  • Create Post
  • heart
    Support Lemmy
  • search
    Search
  • Login
  • Sign Up
LughMA to FuturologyEnglish · 1 year ago

Two-faced AI language models learn to hide deception - ‘Sleeper agents’ seem benign during testing but behave differently once deployed. And methods to stop them aren’t working.

www.nature.com

external-link
message-square
9
link
fedilink
12
external-link

Two-faced AI language models learn to hide deception - ‘Sleeper agents’ seem benign during testing but behave differently once deployed. And methods to stop them aren’t working.

www.nature.com

LughMA to FuturologyEnglish · 1 year ago
message-square
9
link
fedilink
Two-faced AI language models learn to hide deception
www.nature.com
external-link
‘Sleeper agents’ seem benign during testing but behave differently once deployed. And methods to stop them aren’t working.
  • Possibly linux@lemmy.zip
    link
    fedilink
    English
    arrow-up
    1
    ·
    1 year ago

    Sorry, to late for that

    • mateomaui@reddthat.com
      link
      fedilink
      English
      arrow-up
      2
      ·
      1 year ago

      Alright, I’ll be out back digging the bomb shelter.

      • Possibly linux@lemmy.zip
        link
        fedilink
        English
        arrow-up
        1
        ·
        edit-2
        1 year ago

        Its too late for that honestly

        • mateomaui@reddthat.com
          link
          fedilink
          English
          arrow-up
          2
          ·
          1 year ago

          Alright, I’ll switch to digging holes for the family burial ground.

Futurology

futurology

Subscribe from Remote Instance

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: !futurology@futurology.today
Visibility: Public
globe

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

  • 101 users / day
  • 364 users / week
  • 1.71K users / month
  • 5.96K users / 6 months
  • 91 local subscribers
  • 3.06K subscribers
  • 1.95K Posts
  • 12.2K Comments
  • Modlog
  • mods:
  • voidx
  • Lugh
  • Espiritdescali
  • AwesomeLowlander
  • BE: 0.19.11
  • Modlog
  • Legal
  • Instances
  • Docs
  • Code
  • join-lemmy.org