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

    Neuronpedia

    INDEX
    Explanations

    emphasized sections or formatting indicators within the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    highlighted text segments

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    AndEndTag
    -0.86
     wireType
    -0.79
     estekak
    -0.78
    <unused41>
    -0.78
    <unused68>
    -0.78
    [@BOS@]
    -0.78
    <unused51>
    -0.78
    <unused79>
    -0.78
    <unused3>
    -0.77
    <unused8>
    -0.77
    POSITIVE LOGITS
     emphasis
    0.30
    <em>
    0.30
    WillAppear
    0.29
    et
    0.29
     key
    0.28
    underbrace
    0.28
    <b>
    0.27
    item
    0.27
      
    0.26
    numeros
    0.26
    Act Density 0.325%

    No Known Activations