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

    Neuronpedia

    INDEX
    Explanations

    legal documents

    np_max-act · gemini-2.0-flash

    This neuron detects legal‐case party role labels (e.g. “Plaintiff(s),” “Defendant(s),” “Appellant(s),” “Appellee(s)”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ']].
    -0.06
    antis
    -0.06
     order
    -0.06
     hull
    -0.06
     rivalry
    -0.06
     sister
    -0.06
     purity
    -0.06
    RAP
    -0.06
     _.
    -0.06
     acc
    -0.06
    POSITIVE LOGITS
    芸
    0.08
    Scores
    0.07
    鳴
    0.07
    .getHours
    0.07
    าญ
    0.06
    _IMETHOD
    0.06
     marked
    0.06
    Map
    0.06
    移动
    0.06
    _STATE
    0.06
    Act Density 0.004%

    No Known Activations