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

    Neuronpedia

    INDEX
    Explanations

    spokesman

    np_max-act · gemini-2.0-flash

    The neuron detects spokesperson‐attribution phrases introducing quotes (e.g. “A spokesman for X said…”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     ("
    -0.07
    *log
    -0.07
    ival
    -0.06
    ()(
    -0.06
    ')}
    -0.06
     Toni
    -0.06
    }'↵
    -0.06
    (second
    -0.06
    Retrieve
    -0.06
    cert
    -0.06
    POSITIVE LOGITS
     spokesman
    0.11
     spokesperson
    0.09
     spokeswoman
    0.09
    okane
    0.07
    ٪
    0.07
     голос
    0.07
     expressly
    0.07
    .Xr
    0.06
     مي
    0.06
     intra
    0.06
    Act Density 0.004%

    No Known Activations