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

    Neuronpedia

    INDEX
    Explanations

    subjects followed by verbs

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     probably
    0.62
    Probably
    0.58
     Probably
    0.57
    probably
    0.56
     meestal
    0.56
     thường
    0.55
    อาจ
    0.54
    往往
    0.54
     probabilmente
    0.54
    Usually
    0.53
    POSITIVE LOGITS
     ever
    1.02
     fizer
    0.79
     EVER
    0.77
     somehow
    0.75
     fossero
    0.71
     truly
    0.70
     were
    0.69
     forem
    0.68
     tivesse
    0.65
     ooit
    0.63
    Act Density 0.184%

    No Known Activations