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

    Neuronpedia

    INDEX
    Explanations

    qualitative descriptions or characters

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    干燥
    0.39
    čna
    0.37
     drier
    0.37
    popupTheme
    0.36
    тна
    0.35
    heiz
    0.35
    Qui
    0.35
    raina
    0.35
    hack
    0.34
    ánea
    0.34
    POSITIVE LOGITS
     qualitative
    0.57
     symbols
    0.46
     Qualitative
    0.46
     simbol
    0.46
     multid
    0.44
     qualitatively
    0.44
     qualit
    0.44
     parameterized
    0.43
     unstructured
    0.42
     variables
    0.42
    Act Density 0.188%

    No Known Activations