INDEX
    Explanations

    This neuron activates on promotional phrases for online dating or hookup services (e.g. “chat,” “meet singles,” “free dating site”).

    New Auto-Interp
    Negative Logits
    -0.08
    ِك
    -0.07
    工程
    -0.06
    说的
    -0.06
    EAR
    -0.06
    _FILTER
    -0.06
     Emp
    -0.06
     seront
    -0.06
    ?=
    -0.06
    โอ
    -0.06
    POSITIVE LOGITS
     "*
    0.07
     extingu
    0.07
     RHS
    0.07
    (Arrays
    0.07
    0.07
     suffered
    0.07
    qm
    0.06
    bay
    0.06
     svn
    0.06
     ValidationResult
    0.06
    Act Density 0.016%

    No Known Activations