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

    Neuronpedia

    INDEX
    Explanations

    transactions and grammar rules

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     niemals
    0.87
     تمامی
    0.80
     সকল
    0.71
    全く
    0.70
     પોતાના
    0.70
    footage
    0.69
     veramente
    0.67
     nahezu
    0.67
     znacznie
    0.67
     Menschen
    0.66
    POSITIVE LOGITS
     complicating
    0.80
     combinatorial
    0.77
     tricky
    0.74
     reconciling
    0.74
     partly
    0.73
     manipulations
    0.71
     interplay
    0.70
     juggling
    0.70
     rules
    0.68
    紆
    0.68
    Act Density 0.515%

    No Known Activations