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

    Neuronpedia

    INDEX
    Explanations

    text relating to the concept of creating or putting something together as a representation or a model ("mock-up")

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "up."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     needy
    -0.66
     Fidel
    -0.64
     [|
    -0.62
     Boxing
    -0.62
     MAS
    -0.61
    ________________________
    -0.61
     heed
    -0.60
     Ruler
    -0.59
     �
    -0.59
     Sturgeon
    -0.58
    POSITIVE LOGITS
    dates
    1.42
    stairs
    1.20
    dating
    1.17
    olicy
    1.04
    grades
    0.99
    rights
    0.91
    rison
    0.86
    graded
    0.85
    ester
    0.85
    etric
    0.83
    Act Density 0.024%

    No Known Activations