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

    Neuronpedia

    INDEX
    Explanations

    configuration parameters and keywords

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     both
    -1.48
     more
    -1.44
     first
    -1.42
     primarily
    -1.40
    ,
    -1.37
     just
    -1.36
     no
    -1.35
     at
    -1.35
     not
    -1.34
     although
    -1.34
    POSITIVE LOGITS
     colorear
    1.77
    Élet
    1.57
    Vå
    1.55
     chyb
    1.52
     jednod
    1.50
     dénon
    1.48
    pantalones
    1.47
     samoz
    1.46
    هدف
    1.45
    Leírás
    1.45
    Act Density 0.242%

    No Known Activations