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

    Neuronpedia

    INDEX
    Explanations

    fire

    np_max-act · gemini-2.0-flash

    This neuron selectively activates on Portuguese terms for fires or burnings (e.g., words like “incêndio” or “queimadas urbanas”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    озв
    -0.07
     sulfate
    -0.07
    ПО
    -0.07
     debit
    -0.06
    biology
    -0.06
    nicos
    -0.06
     soft
    -0.06
    IGNORE
    -0.06
     berk
    -0.06
    hos
    -0.06
    POSITIVE LOGITS
    蘭
    0.07
    .bias
    0.07
     Fire
    0.07
     wildfires
    0.06
     Farrell
    0.06
    )_
    0.06
     طلا
    0.06
     multitude
    0.06
     війни
    0.06
    TB
    0.06
    Act Density 0.008%

    No Known Activations