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

    Neuronpedia

    INDEX
    Explanations

    recommendation is not

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     even
    -0.84
     their
    -0.81
    moiselle
    -0.80
     karbon
    -0.79
    晰
    -0.76
     every
    -0.74
    <0x99>
    -0.73
     After
    -0.73
     dikutip
    -0.73
     cobre
    -0.73
    POSITIVE LOGITS
     Param
    0.87
    zany
    0.85
     adequ
    0.84
    有些人
    0.82
     certamente
    0.82
    Norma
    0.82
     subst
    0.80
     cq
    0.80
     consequ
    0.79
     ARY
    0.78
    Act Density 0.001%

    No Known Activations