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

    Neuronpedia

    INDEX
    Explanations

    sections of research articles and reports, especially those related to methods and ethics statements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    section separators

    np_acts-logits-general · gemini-2.5-flash-lite

    the neuron detects document structure markers and section headings (e.g., Methods, Data sources, Ethics, Availability) and other formatting/metadata lines.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    nodoc
    -0.57
    ContentAsync
    -0.57
     AssemblyVersion
    -0.52
    bkz
    -0.52
    Gön
    -0.50
    synthesize
    -0.49
    INSEE
    -0.49
    xase
    -0.48
    resco
    -0.48
    tsy
    -0.47
    POSITIVE LOGITS
     natten
    0.40
    AddTagHelper
    0.40
    Dear
    0.40
     mecánico
    0.39
     ética
    0.39
    ьаж
    0.39
     pouvoit
    0.39
     capacitación
    0.39
    TypedDataSet
    0.38
     feroit
    0.37
    Act Density 0.014%

    No Known Activations