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

    Neuronpedia

    INDEX
    Explanations

    moving beyond or towards

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     disappear
    0.89
     affecting
    0.82
     affect
    0.69
     appear
    0.69
     incoming
    0.69
     Appear
    0.69
     vanish
    0.68
     appearing
    0.67
     disappearing
    0.67
     no
    0.66
    POSITIVE LOGITS
     towards
    1.42
     toward
    1.39
    closer
    1.37
    towards
    1.37
    Towards
    1.29
     closer
    1.29
     away
    1.25
     Towards
    1.21
     Toward
    1.17
    Toward
    1.15
    Act Density 0.215%

    No Known Activations