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

    Neuronpedia

    INDEX
    Explanations

    room quality, type, number

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    '
    0.85
    ’
    0.69
    ق
    0.59
    ج
    0.59
    centre
    0.58
    c
    0.58
    brit
    0.57
    bentuk
    0.56
    ۔
    0.56
    argout
    0.55
    POSITIVE LOGITS
    다
    0.64
     ROOM
    0.61
    ло
    0.59
    <0xA3>
    0.56
    ppling
    0.55
    <0x98>
    0.54
     room
    0.54
    ed
    0.54
    తలు
    0.51
    ាតុ
    0.50
    Act Density 0.003%

    No Known Activations