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

    Neuronpedia

    INDEX
    Explanations

    attends to tokens identified as figures or references from tokens marked with specific formatting or coding elements

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.13
    1:0.16
    2:0.12
    3:0.11
    4:0.14
    5:0.15
    6:0.07
    7:0.09
    Negative Logits
     تانيه
    -0.36
    ⓧ
    -0.36
    expandindo
    -0.33
     дописавши
    -0.33
     onAnimation
    -0.33
    matchCondition
    -0.33
    GenerationType
    -0.33
    Revenir
    -0.32
     Administrativna
    -0.32
    Према
    -0.32
    POSITIVE LOGITS
     مشين
    0.24
     pin
    0.23
    blobs
    0.23
    aculture
    0.23
     quedo
    0.22
     Barron
    0.22
    INKS
    0.22
    zzarella
    0.21
    sky
    0.21
     Nebula
    0.21
    Act Density 0.007%

    No Known Activations