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

    Neuronpedia

    INDEX
    Explanations

    ellipsis or placeholder

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     examples
    0.46
     use
    0.44
     example
    0.44
     uses
    0.43
     indicates
    0.42
    /
    0.40
     podemos
    0.40
    我们可以
    0.39
     facilitates
    0.39
     provides
    0.39
    POSITIVE LOGITS
     parecía
    0.41
     lágrimas
    0.41
     తన
    0.40
     había
    0.38
     habían
    0.38
     Gefühl
    0.37
     muttered
    0.37
     hatte
    0.36
     seemed
    0.36
     Wasn
    0.36
    Act Density 0.934%

    No Known Activations