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

    Neuronpedia

    INDEX
    Explanations

    phone components

    np_max-act · gemini-2.0-flash

    The neuron activates on structural/metadata tokens (e.g. header markers) and URL/domain-name fragments in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    �
    -0.07
     avant
    -0.07
     نسخه
    -0.06
    _Delay
    -0.06
    zheimer
    -0.06
     Logo
    -0.06
     thé
    -0.06
     reklam
    -0.06
    posure
    -0.06
    Collect
    -0.06
    POSITIVE LOGITS
     selves
    0.06
     Billing
    0.06
     Dangerous
    0.06
     appeals
    0.06
     Pablo
    0.06
     {!!
    0.06
    ipy
    0.06
     convincing
    0.06
     demol
    0.06
    будь
    0.06
    Act Density 0.026%

    No Known Activations