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

    Neuronpedia

    INDEX
    Explanations

    Machine learning loss functions

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     Braves
    -0.08
    Entered
    -0.08
     í
    -0.08
     cave
    -0.07
     chimp
    -0.07
     MATRIX
    -0.07
     conspir
    -0.07
    utsu
    -0.07
    capitalize
    -0.07
     Ont
    -0.07
    POSITIVE LOGITS
    _loss
    0.11
    Loss
    0.10
    loss
    0.10
    (loss
    0.10
     loss
    0.10
    Combine
    0.09
    -loss
    0.09
     pérdida
    0.09
    .loss
    0.09
     Loss
    0.09
    Act Density 0.004%

    No Known Activations