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

    Neuronpedia

    INDEX
    Explanations

    Denials or accusations

    np_max-act · gemini-2.0-flash

    This neuron detects reporting-style verbs and attribution cues (e.g. said, denied, according to, reported).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Bài
    -0.06
    guna
    -0.06
    ("./
    -0.06
     ATA
    -0.06
    Это
    -0.06
     труд
    -0.06
    ่าว
    -0.06
     Systems
    -0.06
     decrease
    -0.06
     handler
    -0.06
    POSITIVE LOGITS
    іч
    0.07
    Instant
    0.07
    duino
    0.07
    nič
    0.07
    /documentation
    0.06
     rua
    0.06
    мор
    0.06
     TOP
    0.06
    _foot
    0.06
     TP
    0.06
    Act Density 0.024%

    No Known Activations