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

    Neuronpedia

    INDEX
    Explanations

    loss

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     marx
    0.69
     tendencies
    0.67
    Magnet
    0.62
     gun
    0.62
     statically
    0.61
     gunung
    0.60
     bullets
    0.60
    astava
    0.60
     चौराहे
    0.59
     inclined
    0.59
    POSITIVE LOGITS
     loss
    2.42
     lost
    2.34
     losses
    2.31
     Loss
    2.24
    损失
    2.18
    Loss
    2.13
     Lost
    2.08
     Losses
    2.04
    損失
    1.96
    loss
    1.95
    Act Density 0.789%

    No Known Activations