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

    Neuronpedia

    INDEX
    Explanations

    doing whatever you want

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    允许
    0.39
     adequately
    0.38
     urgently
    0.38
     persönlich
    0.38
     любую
    0.38
     appropriately
    0.38
     любой
    0.37
     permettre
    0.37
    Allows
    0.37
     permit
    0.37
    POSITIVE LOGITS
    doing
    0.49
     робити
    0.48
     roam
    0.47
     做
    0.47
    Doing
    0.44
    indon
    0.42
     doing
    0.42
     Doing
    0.42
     whoever
    0.41
     Whatever
    0.40
    Act Density 0.030%

    No Known Activations