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

    Neuronpedia

    INDEX
    Explanations

    assistant

    np_max-act · gemini-2.0-flash

    assistant-role headers/markers indicating the start of an assistant message in chat-formatted text.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     finished
    -0.06
    _episode
    -0.06
    Mir
    -0.06
     měsíců
    -0.06
    ()],
    -0.06
     '');
    -0.06
     metres
    -0.06
    'R
    -0.06
    =text
    -0.06
     Ч
    -0.06
    POSITIVE LOGITS
    عال
    0.06
     incom
    0.06
     ثلاث
    0.06
    -n
    0.06
    ISTIC
    0.06
     آزاد
    0.06
     Jing
    0.06
    での
    0.06
    ног
    0.06
    เขา
    0.06
    Act Density 0.233%

    No Known Activations