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

    Neuronpedia

    INDEX
    Explanations

    optimizer configured

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    份额
    0.47
     valore
    0.45
     الوزن
    0.45
     valeur
    0.44
     initiale
    0.43
     animale
    0.42
     pueda
    0.41
     individuales
    0.41
     œuvre
    0.41
     singoli
    0.40
    POSITIVE LOGITS
    博
    0.42
    CM
    0.42
    ECM
    0.42
    bracketing
    0.41
    ic
    0.41
    CI
    0.41
    SIX
    0.40
    AS
    0.40
    BOARD
    0.39
    SELECT
    0.39
    Act Density 0.003%

    No Known Activations