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

    Neuronpedia

    INDEX
    Explanations

    reference codes, patches, and buttons

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     @
    0.79
    @
    0.79
    Bob
    0.78
    Robert
    0.78
    Harold
    0.76
    <i>
    0.73
    David
    0.73
     Robert
    0.72
    Stanford
    0.70
     rather
    0.70
    POSITIVE LOGITS
    btns
    1.18
     hitbox
    1.03
    btn
    0.98
    闥
    0.98
    adihi
    0.98
    riamo
    0.94
    ार्‍या
    0.94
    button
    0.94
    ванный
    0.94
    ್ರೀ
    0.94
    Act Density 0.001%

    No Known Activations