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

    Neuronpedia

    INDEX
    Explanations

    attends to clauses containing the word "which" from preceding phrases that provide context on various topics

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.06
    1:0.08
    2:0.07
    3:0.12
    4:0.13
    5:0.03
    6:0.33
    7:0.14
    Negative Logits
    acaktır
    -0.29
    our
    -0.27
     our
    -0.27
     Ф
    -0.27
    ten
    -0.26
    ])));
    -0.26
     He
    -0.26
     Our
    -0.25
    eleste
    -0.25
    !"
    -0.25
    POSITIVE LOGITS
    ⓧ
    0.49
     незавершена
    0.49
     autorytatywna
    0.48
    GEBURTSDATUM
    0.48
    ValueStyle
    0.48
     كومونز
    0.48
    Autoritní
    0.45
     beginnetje
    0.44
    WriteTagHelper
    0.44
    JspWriter
    0.44
    Act Density 0.578%

    No Known Activations