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

    Neuronpedia

    INDEX
    Explanations

    what to do or know

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Were
    2.07
    Were
    2.04
    Are
    2.00
     did
    2.00
     are
    1.98
     Are
    1.94
     were
    1.92
    Did
    1.91
     Did
    1.91
     do
    1.86
    POSITIVE LOGITS
     those
    0.88
    those
    0.83
     pitfalls
    0.73
     resonates
    0.70
     each
    0.70
     goodies
    0.67
     maneuvers
    0.67
     succeeds
    0.65
     needs
    0.65
    ₂+
    0.64
    Act Density 0.134%

    No Known Activations