INDEX
    Explanations

    The neuron selectively activates on the specific food item being mentioned in questions or answers about whether a pet can eat it.

    New Auto-Interp
    Negative Logits
     Payload
    -0.07
    后的
    -0.07
    .fp
    -0.06
     вы
    -0.06
    ToMany
    -0.06
    Nickname
    -0.06
    _cod
    -0.06
    (configuration
    -0.06
    १९
    -0.06
    -0.06
    POSITIVE LOGITS
    μέ
    0.07
    soon
    0.07
    вався
    0.06
    sheets
    0.06
    lj
    0.06
    	Task
    0.06
     proportional
    0.06
    0.06
     subordinate
    0.06
     prof
    0.06
    Act Density 0.011%

    No Known Activations