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

    Neuronpedia

    INDEX
    Explanations

    Images and HTML image references

    np_max-act · gemini-2.0-flash

    The neuron is a detector for image‐related references—words like “picture,” “alt,” “text,” “link,” and other tokens used when referring to images or HTML image attributes.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Fur
    -0.06
     laid
    -0.06
     targets
    -0.06
     walked
    -0.06
    .notification
    -0.06
    建筑
    -0.06
    .reddit
    -0.06
     oil
    -0.06
    船
    -0.06
     物
    -0.06
    POSITIVE LOGITS
     üzerinden
    0.07
     порушення
    0.07
    ivol
    0.06
     rejuven
    0.06
    (toolbar
    0.06
    __':
↵
    0.06
     Baron
    0.06
    ▍
    0.06
     emailAddress
    0.06
     "]
    0.06
    Act Density 0.066%

    No Known Activations