Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    phrases related to reversing a decision or position

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to changing one's position or opinion

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ãĥ¼ãĥĨãĤ£
    -0.70
    regate
    -0.68
    oven
    -0.67
    nesota
    -0.67
    unique
    -0.65
    agues
    -0.65
    iciency
    -0.64
    azel
    -0.64
    CLUD
    -0.63
    anon
    -0.63
    POSITIVE LOGITS
     apology
    0.84
     stance
    0.83
     apologizing
    0.82
     decisively
    0.78
     apologise
    0.78
     disav
    0.76
     pledge
    0.75
     withdrawals
    0.75
     pledges
    0.75
     apologies
    0.75
    Act Density 0.238%

    No Known Activations