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

    Neuronpedia

    INDEX
    Explanations

    attends to tokens that express enjoyment or positive sentiment from any later tokens that frame a question or seek clarification

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.44
    1:0.19
    2:0.08
    3:0.06
    4:0.05
    5:0.03
    6:0.05
    7:0.06
    Negative Logits
    Viitteet
    -0.43
    parsedMessage
    -0.39
    HasForeignKey
    -0.37
     "..\..\
    -0.36
    はじめに
    -0.35
    — 
    -0.35
    __(/*!
    -0.35
    ieteur
    -0.35
     &___
    -0.34
    endphp
    -0.34
    POSITIVE LOGITS
    contentLoaded
    0.32
    =$?
    0.31
    ineno
    0.31
    RunWith
    0.30
    Autoritní
    0.30
     méta
    0.29
     surla
    0.29
     Geruch
    0.28
    DispatchToProps
    0.28
     émotion
    0.28
    Act Density 0.531%

    No Known Activations