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

    Neuronpedia

    INDEX
    Explanations

    titles or headings marked with a special character sequence

    oai_token-act-pair · gpt-3.5-turbo

    titles of articles, reports, or creative works

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     range
    -0.83
    theless
    -0.82
     infring
    -0.77
     proportion
    -0.73
     sear
    -0.72
     Belg
    -0.69
     regardless
    -0.67
     fraction
    -0.66
     anyway
    -0.65
     bombard
    -0.65
    POSITIVE LOGITS
    ª
    1.63
    ł
    1.46
    ı
    1.44
    ³
    1.36
    «
    1.35
    Ĵ
    1.35
    ¹
    1.35
    §
    1.32
    IJ
    1.30
    Ķ
    1.28
    Act Density 0.139%

    No Known Activations