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

    Neuronpedia

    INDEX
    Explanations

    Hol-" prefixes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     agreements
    -0.78
     widely
    -0.76
    ogos
    -0.73
    勤
    -0.72
     rocks
    -0.72
    mặt
    -0.72
    ankind
    -0.72
     since
    -0.71
     establishment
    -0.70
     Vern
    -0.69
    POSITIVE LOGITS
     Hol
    1.11
     HOL
    1.08
    Hol
    1.00
     הול
    0.98
     Holstein
    0.95
    thold
    0.92
     هول
    0.90
     hol
    0.90
     mineures
    0.89
     HOLD
    0.84
    Act Density 0.007%

    No Known Activations