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

    Neuronpedia

    INDEX
    Explanations

    safe boundaries

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     namesake
    0.51
     pose
    0.50
     less
    0.47
     response
    0.45
     significant
    0.45
     sustain
    0.45
     typically
    0.45
     major
    0.45
     low
    0.44
     link
    0.43
    POSITIVE LOGITS
     এমনকি
    0.63
     কিংবা
    0.61
     thậm
    0.59
     Bahkan
    0.56
     অথবা
    0.55
     Etc
    0.54
     hatta
    0.53
     Incluso
    0.47
     bahkan
    0.45
     Surely
    0.44
    Act Density 0.252%

    No Known Activations