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

    Neuronpedia

    INDEX
    Explanations

    phrase related to memories or instructions to remember something

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "forget."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ccording
    -0.70
    uana
    -0.67
    ioxide
    -0.66
    uilt
    -0.64
     purported
    -0.63
     Purs
    -0.62
    inction
    -0.62
    ivil
    -0.62
    FO
    -0.62
    Ale
    -0.61
    POSITIVE LOGITS
    fulness
    1.10
    ful
    1.04
     forget
    1.03
    fully
    0.95
     forgot
    0.90
    theless
    0.84
    ãĤ¤ãĥĪ
    0.84
     forgetting
    0.83
    bryce
    0.81
    bolt
    0.78
    Act Density 0.009%

    No Known Activations