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

    Neuronpedia

    INDEX
    Explanations

    technical documents

    np_max-act · gemini-2.0-flash

    This neuron detects Korean phrases in prompts that specify surgical‐procedure details—i.e. instructions about which organ was treated, its precise location, and the procedure identifier.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    пон
    -0.07
     sleek
    -0.07
     decks
    -0.07
    uls
    -0.07
    ẻ
    -0.07
     pulse
    -0.06
    GLOSS
    -0.06
    �
    -0.06
     group
    -0.06
     jaws
    -0.06
    POSITIVE LOGITS
    로는
    0.06
    زی
    0.06
    Justin
    0.06
    .AddWithValue
    0.06
     approached
    0.06
    Sym
    0.06
     him
    0.06
    Created
    0.06
    ьми
    0.06
     Hearth
    0.06
    Act Density 0.005%

    No Known Activations