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

    Neuronpedia

    INDEX
    Explanations

    distinct concepts or issues

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     overwhelmingly
    0.52
    the
    0.51
     underlies
    0.51
     hasn
    0.50
     expects
    0.50
     the
    0.48
    りで
    0.47
     clears
    0.47
    ０
    0.47
     denies
    0.46
    POSITIVE LOGITS
    ل
    0.64
    /
    0.54
    }-
    0.49
    充分
    0.48
    FindingsResponse
    0.48
    晠
    0.47
    }+\
    0.47
    Batch
    0.45
     Mfg
    0.45
    রস
    0.45
    Act Density 0.000%

    No Known Activations