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

    Neuronpedia

    INDEX
    Explanations

    quantifiable descriptors

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     moveTo
    0.49
    ).
    0.49
     cutoff
    0.48
     refill
    0.48
     slings
    0.47
     wipe
    0.47
     preposition
    0.47
    )…
    0.46
     bounce
    0.46
     doen
    0.45
    POSITIVE LOGITS
     рассказа
    0.55
     Осо
    0.54
     музи
    0.54
    ኋ
    0.54
     розпо
    0.53
     же
    0.52
    an
    0.52
    น
    0.52
     ад
    0.52
    ા
    0.52
    Act Density 0.001%

    No Known Activations