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

    Neuronpedia

    INDEX
    Explanations

    programming-related keywords and syntax

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

    protects against

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron detects the start-of-sequence (beginning-of-document) token.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    出版年
    -1.00
     ब्रेकडाउन
    -0.86
     تضيفلها
    -0.74
    KommentareTeilen
    -0.73
    <unused8>
    -0.72
    باردا
    -0.72
    <unused43>
    -0.72
    <unused16>
    -0.72
    <pad>
    -0.72
    [@BOS@]
    -0.71
    POSITIVE LOGITS
    import
    0.52
    <strong>
    0.51
    <h1>
    0.50
    <
    0.47
    <h2>
    0.46
    The
    0.44
    I
    0.42
    ↵↵
    0.42
    [
    0.42
    ↵
    0.42
    Act Density 0.000%

    No Known Activations