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

    Neuronpedia

    INDEX
    Explanations

    outer or lower

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ς
    0.87
     पति
    0.75
     '-':
    0.74
     $*$-
    0.74
    ouflage
    0.74
    اة
    0.74
    籍
    0.73
     '':
    0.73
    Гц
    0.73
    ρές
    0.72
    POSITIVE LOGITS
    er
    1.00
    ER
    0.83
    ﻨ
    0.77
    most
    0.77
    ο
    0.73
     kunde
    0.72
    halb
    0.70
    well
    0.68
    ON
    0.67
    zeige
    0.67
    Act Density 0.063%

    No Known Activations