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

    Neuronpedia

    INDEX
    Explanations

    Q

    np_max-act · gemini-2.0-flash

    This neuron detects the “Q” and “A” speaker‐label markers (i.e. question/answer notations) in transcript excerpts.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Kara
    -0.07
    .Copy
    -0.06
    Hz
    -0.06
     gest
    -0.06
     fails
    -0.06
     minced
    -0.06
     finger
    -0.06
     copying
    -0.06
     meat
    -0.06
    count
    -0.06
    POSITIVE LOGITS
    detail
    0.07
     Asheville
    0.06
    ButtonModule
    0.06
     визнача
    0.06
    ній
    0.06
    _STAR
    0.06
    -relative
    0.06
     upcoming
    0.06
     olmuştur
    0.06
    บาล
    0.06
    Act Density 0.002%

    No Known Activations