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

    Neuronpedia

    INDEX
    Explanations

    before decisions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ular
    0.68
    enen
    0.67
    ins
    0.67
    ,
    0.66
     then
    0.66
    mselves
    0.65
    elf
    0.65
    well
    0.65
     and
    0.64
    ium
    0.64
    POSITIVE LOGITS
     Decisions
    1.25
     Always
    1.16
     If
    1.15
     Các
    1.15
     Entscheid
    1.13
     Everything
    1.12
     Jeśli
    1.10
     Hvis
    1.08
     Everyone
    1.08
     Different
    1.05
    Act Density 0.011%

    No Known Activations