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

    Neuronpedia

    INDEX
    Explanations

    leaving or ending phrases

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     genocide
    0.46
     neben
    0.45
     lotus
    0.45
     Chalmers
    0.44
     yoga
    0.44
     humains
    0.43
     bookkeeping
    0.43
     поряд
    0.43
     персо
    0.42
     Lotus
    0.42
    POSITIVE LOGITS
    NGC
    0.48
    Jeśli
    0.41
    Treas
    0.39
    Prop
    0.38
    Why
    0.37
    why
    0.36
    ુપ
    0.36
    ReLU
    0.36
    What
    0.35
    Would
    0.35
    Act Density 0.000%

    No Known Activations