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

    Neuronpedia

    INDEX
    Explanations

    perhaps elsewhere

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     generally
    0.56
     gets
    0.55
    Additionally
    0.54
     definitely
    0.50
     overall
    0.49
     goof
    0.48
     Additionally
    0.47
     whopping
    0.46
     shenanigans
    0.46
     snooze
    0.46
    POSITIVE LOGITS
     Perhaps
    0.67
    perhaps
    0.66
     Beneath
    0.64
    Perhaps
    0.61
     Tonight
    0.55
     Slowly
    0.55
     Porque
    0.53
    每一个
    0.52
     Somewhere
    0.52
     Mỗi
    0.50
    Act Density 0.137%

    No Known Activations