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

    Neuronpedia

    INDEX
    Explanations

    air

    np_max-act · gemini-2.0-flash

    This neuron detects mentions of air ingress or airflow (e.g., “air,” “enter,” “sucked back into”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     setValue
    -0.08
    (getResources
    -0.07
     fieldValue
    -0.07
     preference
    -0.06
    ССР
    -0.06
     isSuccess
    -0.06
     Goodman
    -0.06
    788
    -0.06
    ()↵↵↵↵
    -0.06
     Müdürlüğü
    -0.06
    POSITIVE LOGITS
    onium
    0.07
     Lingu
    0.07
     climate
    0.07
     тип
    0.07
    μαι
    0.07
     deb
    0.07
    nette
    0.07
     lu
    0.07
     adjoining
    0.06
     apply
    0.06
    Act Density 0.009%

    No Known Activations