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

    Neuronpedia

    INDEX
    Explanations

    说 (to say, to speak)

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    やり
    0.39
     अभि
    0.36
     komunik
    0.36
    害
    0.36
     Expr
    0.36
    શા
    0.36
     comunic
    0.36
    ússia
    0.35
     yath
    0.35
    '}$
    0.35
    POSITIVE LOGITS
    UserDetails
    0.40
     überzeugt
    0.40
    Compact
    0.39
     secrets
    0.39
    Gob
    0.39
    secrets
    0.38
    \}
    0.37
    ="|
    0.37
     sob
    0.37
    Secrets
    0.36
    Act Density 0.003%

    No Known Activations