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

    Neuronpedia

    INDEX
    Explanations

    Numerical groupings

    np_max-act · gemini-2.0-flash

    This neuron detects mentions of numeric counts and list structures—terms that enumerate or index items (e.g., “two,” “four,” “16,” paired with words like “variable,” “conditions,” “model,” or “classifications”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     plagiarism
    -0.07
    ABB
    -0.07
    abb
    -0.06
    computer
    -0.06
     Myers
    -0.06
     mostly
    -0.06
     Psi
    -0.06
     Girlfriend
    -0.06
     Raleigh
    -0.06
    _nonce
    -0.06
    POSITIVE LOGITS
    坦
    0.08
     jung
    0.07
    etrize
    0.06
     ges
    0.06
     forme
    0.06
    think
    0.06
    атегор
    0.06
    丘
    0.06
    	UI
    0.06
     msgid
    0.06
    Act Density 0.096%

    No Known Activations