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

    Neuronpedia

    INDEX
    Explanations

    uc

    np_max-act · gemini-2.0-flash

    This neuron detects occurrences of the term “Confucianism” (its subword fragments like “Conf,” “uc,” “ian,” “ism”) or direct references to that philosophical school.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Starting
    -0.07
     whip
    -0.07
     spiders
    -0.07
    _BROWSER
    -0.07
     {?
    -0.07
    _POS
    -0.06
    AMB
    -0.06
    弹
    -0.06
     مراج
    -0.06
     Heights
    -0.06
    POSITIVE LOGITS
     confidential
    0.07
    control
    0.06
     Assign
    0.06
                                    ↵
    0.06
    iven
    0.06
    hões
    0.06
    bras
    0.06
    apps
    0.06
     convertView
    0.06
    //↵↵↵
    0.06
    Act Density 0.002%

    No Known Activations