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

    Neuronpedia

    INDEX
    Explanations

    direct modifier

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    {
    0.47
     étend
    0.46
    )$
    0.45
     الك
    0.44
    LL
    0.43
     الكل
    0.43
     accél
    0.42
     hinzu
    0.42
    URNS
    0.41
    પ
    0.41
    POSITIVE LOGITS
     direct
    1.30
     directly
    1.29
    directly
    1.26
    直接
    1.25
     bezpośred
    1.25
     Directly
    1.24
     Direct
    1.21
    direct
    1.21
    Direct
    1.20
     напрямую
    1.20
    Act Density 0.035%

    No Known Activations