• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
AI
Report

Praise for an experiment's design and results

A user calls the experiment's design clever and its results “very interesting,” without giving further details.

Dylan HadfieldMenellDH
xuan (ɕɥɛn / sh-yen)X(
Iason GabrielIG
7 Sources, 2h ago, first seen 2h ago

TLDR

In a quote post, a user praises an experiment's design as clever and calls its results “very interesting.” The comment does not say what the experiment tested or found.

Combined views

3.3K

7 Sources, first seen 2h ago

44 likes3 comments29 saves12 reposts

Combined views

3.3K

7 Sources, first seen 2h ago

44 likes3 comments29 saves12 reposts
Image: @PKirgis on X

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

Featured Source

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

7 Sources

Peter Kirgis@PKirgisFrontier AI companies use constitutions, model specifications, and system prompts to guide the moral orientation of their models. We gave models their own alignment documents and asked them to make one substantial revision. Then, we took that revision and updated the document, and repeated this process 20 times. 🧵2h
Seth Lazar@sethlazarThis is a clever experiment design with very interesting results!2h
Iason Gabriel@IasonGabrielFascinating research on what happens when AI models are allowed to edit their own constitutions, including shifts in the parties they favor and the authority relations they endorse:1h
Nick Caputo@nickacaputoSuper cool project! It's great to see questions about how the models think about their own constitutions and what alignment means explored empirically, especially across a trajectory of amendment like this1h
Atoosa Kasirzadeh@Dr_AtoosaMust read and ponder about50m
xuan (ɕɥɛn / sh-yen)@xuanalogueRT @PKirgis: Frontier AI companies use constitutions, model specifications, and system prompts to guide the moral orientation of their mode…45m
Dylan HadfieldMenell@dhadfieldmenellRT @nickacaputo: Super cool project! It's great to see questions about how the models think about their own constitutions and what alignmen…36m
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI

    7 Sources

    Peter Kirgis@PKirgisFrontier AI companies use constitutions, model specifications, and system prompts to guide the moral orientation of their models. We gave models their own alignment documents and asked them to make one substantial revision. Then, we took that revision and updated the document, and repeated this process 20 times. 🧵2h
    Seth Lazar@sethlazarThis is a clever experiment design with very interesting results!2h
    Iason Gabriel@IasonGabrielFascinating research on what happens when AI models are allowed to edit their own constitutions, including shifts in the parties they favor and the authority relations they endorse:1h
    Nick Caputo@nickacaputoSuper cool project! It's great to see questions about how the models think about their own constitutions and what alignment means explored empirically, especially across a trajectory of amendment like this1h
    Atoosa Kasirzadeh@Dr_AtoosaMust read and ponder about50m
    xuan (ɕɥɛn / sh-yen)@xuanalogueRT @PKirgis: Frontier AI companies use constitutions, model specifications, and system prompts to guide the moral orientation of their mode…45m
    Dylan HadfieldMenell@dhadfieldmenellRT @nickacaputo: Super cool project! It's great to see questions about how the models think about their own constitutions and what alignmen…36m
    Today's Rank

    #3

    Today's Rank

    #3