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

    Neuronpedia

    INDEX
    Explanations

    attends to punctuation-related tokens from tokens that mention page numbers or related academic references

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.08
    1:0.10
    2:0.08
    3:0.32
    4:0.16
    5:0.03
    6:0.09
    7:0.10
    Negative Logits
    توض
    -0.35
     bezeichneter
    -0.33
    BorderLayout
    -0.32
    KeySpec
    -0.32
     Waray
    -0.32
     defaultstate
    -0.32
    новништво
    -0.31
    ArgsConstructor
    -0.31
    TabIndex
    -0.31
    StoryboardSegue
    -0.30
    POSITIVE LOGITS
    MLLoader
    0.40
     فريبيس
    0.30
     Chwiliwch
    0.29
    RegistryLite
    0.28
     Carrier
    0.28
     carrier
    0.27
    LTR
    0.27
    carrier
    0.27
    ondissement
    0.27
     ус
    0.27
    Act Density 0.406%

    No Known Activations