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

    Neuronpedia

    INDEX
    Explanations

    Expressing opinions

    np_max-act · gemini-2.0-flash

    The neuron is a general “quantifier/demonstrative” detector—it lights up on words like “any,” “every,” “that,” “the,” “an,” “young,” “book,” etc., i.e. tokens used to make generalizing or specifying statements.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    oscopic
    -0.07
    -independent
    -0.07
     Technical
    -0.06
    ']]]↵
    -0.06
     qb
    -0.06
    OS
    -0.06
    robot
    -0.06
    _Menu
    -0.06
     Organizations
    -0.06
    Terminate
    -0.06
    POSITIVE LOGITS
    .Normalize
    0.06
    ByKey
    0.06
    �
    0.06
     Cemetery
    0.06
    setType
    0.06
    .mapper
    0.06
    _preds
    0.06
    /--
    0.06
    盖
    0.06
     Impl
    0.06
    Act Density 0.094%

    No Known Activations