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

    Neuronpedia

    INDEX
    Explanations

    contrastive loss and objectives

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     saison
    0.66
     florist
    0.66
     olhos
    0.61
     bisc
    0.61
     vyh
    0.60
     ainda
    0.59
     જો
    0.59
     mais
    0.59
     unn
    0.59
     BMW
    0.58
    POSITIVE LOGITS
    ون
    0.57
    를
    0.57
    І
    0.55
    atches
    0.54
    ೀಲ
    0.54
    Assignments
    0.53
    elere
    0.53
    nasium
    0.52
    嚧
    0.52
    gaben
    0.52
    Act Density 0.041%

    No Known Activations