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

    Neuronpedia

    INDEX
    Explanations

    off

    np_max-act · gemini-2.0-flash

    The neuron fires on adverbs and verbs that convey covert or stealthy actions (e.g. “secretly,” “drifted”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    LocalStorage
    -0.07
    ({"
    -0.06
    namen
    -0.06
    .checked
    -0.06
     dép
    -0.06
    _DEPRECATED
    -0.06
    @@@@
    -0.06
     spotting
    -0.06
    'all
    -0.06
    (selected
    -0.06
    POSITIVE LOGITS
     جا
    0.07
    .kr
    0.07
     पढ
    0.07
     فول
    0.06
    ROT
    0.06
     poorly
    0.06
     \↵
    0.06
    EG
    0.06
    FUNC
    0.06
     cheesy
    0.06
    Act Density 0.908%

    No Known Activations