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

    Neuronpedia

    INDEX
    Explanations

    say "stream methods"

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Host
    0.79
     सोना
    0.74
     Host
    0.72
    Eng
    0.70
     Organ
    0.69
    これ
    0.67
     org
    0.67
     host
    0.66
     organ
    0.66
    First
    0.65
    POSITIVE LOGITS
     flat
    0.71
     celebrate
    0.66
    Flat
    0.66
     Flat
    0.65
     celebrar
    0.64
    getic
    0.64
     admit
    0.63
     flatter
    0.63
     ستار
    0.63
     एग्जाम
    0.62
    Act Density 0.007%

    No Known Activations