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

    Neuronpedia

    INDEX
    Explanations

    ie

    np_max-act · gemini-2.0-flash

    This neuron specializes in spotting tokens that occur in infobox image‐file names or captions (e.g. parts of “.jpg”/“.png” filenames and associated image-title strings).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     fins
    -0.07
    	up
    -0.07
    KeyType
    -0.06
     lesbisk
    -0.06
     pur
    -0.06
     elevator
    -0.06
    .addNode
    -0.06
    .Get
    -0.06
     estilo
    -0.06
     野
    -0.06
    POSITIVE LOGITS
     subsidiaries
    0.06
     trí
    0.06
    usize
    0.06
    �
    0.06
    TypeEnum
    0.06
     روم
    0.06
     ас
    0.06
    แหน
    0.06
    าการ
    0.06
    -tw
    0.06
    Act Density 0.055%

    No Known Activations