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

    Neuronpedia

    INDEX
    Explanations

    we need or have

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    据说
    0.57
     reportedly
    0.55
    据悉
    0.54
    presumably
    0.50
    apparently
    0.50
     ternyata
    0.49
     apparently
    0.49
     Apparently
    0.45
    nsp
    0.45
    据
    0.44
    POSITIVE LOGITS
     deserved
    0.59
     deserve
    0.58
     should
    0.56
     Should
    0.56
     deserves
    0.56
     best
    0.55
     devraient
    0.54
     unfairly
    0.53
     mérite
    0.51
     most
    0.51
    Act Density 0.051%

    No Known Activations