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

    Neuronpedia

    INDEX
    Explanations

    nullable data type handling

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    either
    -1.02
    too
    -0.89
     เช
    -0.88
    anything
    -0.88
    âts
    -0.86
    yin
    -0.84
     fi
    -0.77
    ff
    -0.77
     ctypes
    -0.76
    "…
    -0.76
    POSITIVE LOGITS
     ??
    3.33
    ??
    2.55
     ???
    1.96
    ???
    1.94
    ??"
    1.93
     or
    1.90
    ??)
    1.81
    ????
    1.77
     ????
    1.76
     ?????
    1.52
    Act Density 0.026%

    No Known Activations