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

    Neuronpedia

    INDEX
    Explanations

    understand you, break down

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ).
    0.88
     workpiece
    0.87
     verwendet
    0.82
     brukes
    0.81
     lure
    0.80
     puede
    0.79
    ).)
    0.79
     eyepiece
    0.79
    spapers
    0.78
    )")
    0.77
    POSITIVE LOGITS
     the
    0.83
     our
    0.75
     these
    0.74
    पी
    0.74
     τις
    0.74
     this
    0.74
     आम्हाला
    0.73
     Май
    0.73
    増
    0.71
    Пі
    0.71
    Act Density 2.522%

    No Known Activations