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

    Neuronpedia

    INDEX
    Explanations

    instances of punctuation and apostrophes, indicating informal or conversational text

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

    Apostrophes, periods, or "eta" followed by letters

    np_acts-logits-general · gemini-2.0-flash

    possessive 's or closing quotes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     للاسماء
    -0.81
     مشين
    -0.70
     ***!
    -0.64
    UnsafeEnabled
    -0.62
    )";
    -0.62
    RetentionPolicy
    -0.61
     الحره
    -0.60
    `).
    -0.60
    BorderFactory
    -0.59
    DockStyle
    -0.59
    POSITIVE LOGITS
    “
    0.59
     sherds
    0.56
     sputnik
    0.54
     Oedipus
    0.53
    <em>
    0.51
    Jîn
    0.51
     authorship
    0.51
     stoneware
    0.50
    s
    0.50
     Sodom
    0.50
    Act Density 0.076%

    No Known Activations