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

    Neuronpedia

    INDEX
    Explanations

    addressing the reader

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     does
    1.15
    does
    0.96
    Does
    0.95
     Does
    0.90
     doesn
    0.84
     did
    0.81
     is
    0.75
     do
    0.75
    doesn
    0.74
    Might
    0.74
    POSITIVE LOGITS
     ਤੁਸੀਂ
    0.91
     he
    0.85
     they
    0.83
     theyre
    0.83
     он
    0.80
     তাঁরা
    0.76
     그는
    0.76
    他就
    0.75
     THEY
    0.74
     они
    0.73
    Act Density 0.409%

    No Known Activations