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

    Neuronpedia

    INDEX
    Explanations

    closely

    np_max-act · gemini-2.0-flash

    This neuron detects adverbs (especially “closely,” “deeply,” etc.) used to emphasize or qualify relationships between concepts.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    rve
    -0.07
    572
    -0.06
    truck
    -0.06
    (Property
    -0.06
    Upgrade
    -0.06
    .Var
    -0.06
     MyApp
    -0.06
    gpu
    -0.06
    arlar
    -0.06
    jh
    -0.06
    POSITIVE LOGITS
     closely
    0.10
    avoid
    0.07
     examining
    0.07
     міг
    0.07
     constantly
    0.07
     тобто
    0.07
     membership
    0.07
    BMI
    0.07
     tình
    0.06
     examine
    0.06
    Act Density 0.012%

    No Known Activations