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

    Neuronpedia

    INDEX
    Explanations

    actions to guide behavior

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     جميع
    0.51
    isins
    0.49
    جميع
    0.47
     اپنی
    0.44
     vanwege
    0.44
    oplane
    0.43
     perfettamente
    0.43
     segala
    0.43
     completo
    0.41
     allt
    0.40
    POSITIVE LOGITS
     ώστε
    0.60
     selectively
    0.58
     incentivize
    0.55
     penal
    0.54
     allow
    0.52
     incentiv
    0.50
     proactively
    0.50
     zwięks
    0.50
     ensure
    0.49
     discourage
    0.49
    Act Density 0.105%

    No Known Activations