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

    Neuronpedia

    INDEX
    Explanations

    first person

    np_max-act · gemini-2.0-flash

    This neuron primarily detects self‐referential language, especially first‐person pronouns (e.g. “I,” “me,” “my”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Containing
    -0.07
    descending
    -0.06
    _summary
    -0.06
    secutive
    -0.06
    Util
    -0.06
    keyboard
    -0.06
    _sink
    -0.06
    datasets
    -0.06
     compass
    -0.06
    -ring
    -0.06
    POSITIVE LOGITS
     suoi
    0.07
     titre
    0.06
    closest
    0.06
     Prevent
    0.06
     ATF
    0.06
    ibr
    0.06
    esor
    0.06
    .est
    0.06
     paycheck
    0.06
     TX
    0.06
    Act Density 0.060%

    No Known Activations