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

    Neuronpedia

    INDEX
    Explanations

    I am programmed to refuse

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Jeffrey
    0.42
    द्वीप
    0.40
     Legislative
    0.38
    pans
    0.36
     counters
    0.36
     Jeffrey
    0.36
     spans
    0.35
    Anita
    0.35
     திட்ட
    0.35
    ❒
    0.35
    POSITIVE LOGITS
     facie
    0.39
     디지털
    0.38
     ranked
    0.38
     ống
    0.36
    Merc
    0.35
     manus
    0.35
    rí
    0.35
    远程
    0.35
    排
    0.35
    र्ग
    0.35
    Act Density 0.009%

    No Known Activations