INDEX
    Explanations

    The main thing this neuron does is detect mentions of shipping costs or shipping‐related details (e.g. “shipping” followed by a price or rate).

    New Auto-Interp
    Negative Logits
     usted
    -0.07
    zano
    -0.06
     Muss
    -0.06
    -0.06
    Phones
    -0.06
    ��
    -0.06
    vid
    -0.06
    -0.06
    ces
    -0.06
    Monitor
    -0.06
    POSITIVE LOGITS
     shipping
    0.08
    abstract
    0.07
    ;.
    0.07
    github
    0.07
    _KIND
    0.07
     }
    ↵
    0.06
    (reply
    0.06
     recognition
    0.06
     cowork
    0.06
     start
    0.06
    Act Density 0.006%

    No Known Activations