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

    Neuronpedia

    INDEX
    Explanations

    ask, give, interact

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     attachments
    0.45
     pleads
    0.45
     vows
    0.43
     rotting
    0.43
     vomit
    0.41
     obscene
    0.41
     attachment
    0.40
     retails
    0.40
     مرف
    0.40
     rotten
    0.39
    POSITIVE LOGITS
     accusing
    0.68
    考验
    0.60
     accuse
    0.60
    给我
    0.57
     என்னை
    0.55
    給我
    0.54
    আমাকে
    0.54
     interacting
    0.53
    把我
    0.53
     treating
    0.52
    Act Density 0.058%

    No Known Activations