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

    Neuronpedia

    INDEX
    Explanations

    fantastic / excellent / loved / wow

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     exist
    0.38
     特定
    0.37
     tempt
    0.37
     embrace
    0.36
     شوند
    0.35
     dangle
    0.35
     exists
    0.35
    FilesIn
    0.35
    どのように
    0.35
    ள்
    0.34
    POSITIVE LOGITS
    Absolutely
    0.93
    Wow
    0.87
    Fantastic
    0.84
    Loved
    0.84
    absolutely
    0.84
    Excellent
    0.83
    Really
    0.83
     Absolutely
    0.82
    Excelente
    0.79
    Honestly
    0.79
    Act Density 0.014%

    No Known Activations