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

    Neuronpedia

    INDEX
    Explanations

    problem and answer structure

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    .")]
    0.76
    .},
    0.72
    .}\
    0.70
     spaceShip
    0.69
    巌
    0.69
    }/>
    0.69
     perfil
    0.69
    ংকর
    0.68
    .');
    0.68
    issory
    0.68
    POSITIVE LOGITS
    While
    1.61
     Although
    1.56
     While
    1.56
    Although
    1.54
    Despite
    1.50
    Since
    1.47
     Despite
    1.46
    When
    1.44
     Unlike
    1.38
    This
    1.38
    Act Density 0.398%

    No Known Activations