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

    Neuronpedia

    INDEX
    Explanations

    instances of contractions and possessive forms in text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ſelf
    -1.33
     iſt
    -1.25
     Efq
    -1.21
    ſelves
    -1.21
     itſelf
    -1.20
     ―――――
    -1.18
     ་་
    -1.17
     pleaſure
    -1.17
     uſ
    -1.17
     uſe
    -1.16
    POSITIVE LOGITS
     
    0.85
     I
    0.81
    ↵↵
    0.73
    ↵
    0.70
     it
    0.67
    !
    0.67
    <eos>
    0.64
    ..
    0.64
    .
    0.63
     x
    0.63
    Act Density 0.228%

    No Known Activations