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

    Neuronpedia

    INDEX
    Explanations

    water/sea

    np_max-act · gemini-2.0-flash

    The neuron activates on domain‐type adjectives describing environments or habitats—especially water‐ and air‐related terms like coastal, marine/maritime, aquatic, aerial, waterborne, etc.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     punt
    -0.07
     soy
    -0.06
     стан
    -0.06
     Von
    -0.06
    Cold
    -0.06
     unfinished
    -0.06
    ़ो
    -0.06
     wounded
    -0.06
     shield
    -0.06
    house
    -0.06
    POSITIVE LOGITS
    letion
    0.07
    bsub
    0.07
     DOM
    0.06
     вт
    0.06
    (tags
    0.06
     수상
    0.06
     ThemeData
    0.06
    aut
    0.06
     intersection
    0.06
    oại
    0.06
    Act Density 0.017%

    No Known Activations