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

    Neuronpedia

    INDEX
    Explanations

    tech issues

    np_max-act · gemini-2.0-flash

    The neuron activates on words that describe impact or interference (e.g. “impede,” “affect”) on user experience or performance.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     projected
    -0.07
    .sms
    -0.07
     مرکزی
    -0.07
    ongo
    -0.06
     sauces
    -0.06
     SIP
    -0.06
    words
    -0.06
    sembles
    -0.06
    UTURE
    -0.06
    σιο
    -0.06
    POSITIVE LOGITS
    ،↵
    0.07
    >',↵
    0.06
    babel
    0.06
     tearing
    0.06
     ét
    0.06
     animations
    0.06
    _metadata
    0.06
     इतन
    0.06
    +='<
    0.06
     „
    0.06
    Act Density 0.031%

    No Known Activations