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

    Neuronpedia

    INDEX
    Explanations

    have learned or been chosen

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Had
    0.46
     HAD
    0.42
    Had
    0.42
    took
    0.41
    casters
    0.40
    had
    0.39
     hadrons
    0.37
    pris
    0.36
    aitement
    0.36
    HAD
    0.35
    POSITIVE LOGITS
     been
    0.61
     bisher
    0.51
     previously
    0.49
     chosen
    0.49
     scoped
    0.47
     lopen
    0.46
     begun
    0.46
     liggen
    0.46
     précédemment
    0.45
     already
    0.45
    Act Density 0.021%

    No Known Activations