INDEX
    Explanations

    The neuron detects compound modifier phrases labeling types of factors—especially hyphenated descriptors ending in “-related” (e.g. injury-related, disease-related) or demographic descriptors.

    New Auto-Interp
    Negative Logits
     диза
    -0.07
    .getLast
    -0.06
     walnut
    -0.06
     lite
    -0.06
    -0.05
    Роз
    -0.05
    />.↵↵
    -0.05
     Bi
    -0.05
    ERVICE
    -0.05
     Toe
    -0.05
    POSITIVE LOGITS
    (cl
    0.08
     peptides
    0.07
    lyph
    0.07
    arity
    0.07
     음악
    0.06
     Reads
    0.06
    xlsx
    0.06
     Additionally
    0.06
     repayment
    0.06
    0.06
    Act Density 0.052%

    No Known Activations