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

    Neuronpedia

    INDEX
    Explanations

    changing or not changing

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     lediglich
    0.63
     Incidentally
    0.60
     aptly
    0.54
    Incidentally
    0.52
     partake
    0.51
     intricately
    0.51
     expts
    0.50
     लिहाजा
    0.50
     accolades
    0.50
     hence
    0.49
    POSITIVE LOGITS
     beside
    0.65
    maybe
    0.61
     informations
    0.61
     standart
    0.60
     Beside
    0.60
     drawed
    0.59
     really
    0.59
     witch
    0.59
     maybe
    0.57
     Maybe
    0.56
    Act Density 0.010%

    No Known Activations