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

    Neuronpedia

    INDEX
    Explanations

    Uncommon words

    np_max-act · gemini-2.0-flash

    This neuron fires on descriptive appearance modifiers—words that characterize surface quality or visual style (e.g. glimmering, film-like, handsome).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     chewing
    -0.07
     kans
    -0.07
    900
    -0.06
     personas
    -0.06
    untary
    -0.06
    iou
    -0.06
     mak
    -0.06
     hog
    -0.06
    ucci
    -0.06
    ang
    -0.06
    POSITIVE LOGITS
    `='$
    0.08
     getUsername
    0.07
    "?>↵
    0.06
    trieve
    0.06
    _Request
    0.06
     MEP
    0.06
    SOEVER
    0.06
    .Flat
    0.06
     مذ
    0.06
     irresist
    0.06
    Act Density 0.588%

    No Known Activations