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

    Neuronpedia

    INDEX
    Explanations

    set each factor to zero

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ag
    1.18
    ً
    1.13
    ো
    1.11
     remaining
    1.05
    ir
    1.02
    os
    1.01
    on
    1.01
    h
    1.00
     flip
    0.99
    j
    0.98
    POSITIVE LOGITS
    vegarde
    1.30
     thei
    1.23
    datei
    1.22
    cusson
    1.20
    áculo
    1.18
    ែល
    1.17
    vorschau
    1.17
    sette
    1.14
    date
    1.14
    outube
    1.14
    Act Density 0.002%

    No Known Activations