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

    Neuronpedia

    INDEX
    Explanations

    section

    np_max-act · gemini-2.0-flash

    This neuron detects occurrences of the word “section,” as used when citing numbered sections (e.g., of laws or documents).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    AMY
    -0.08
    696
    -0.07
    amy
    -0.07
     Memory
    -0.07
    (wallet
    -0.07
     glm
    -0.07
    Fly
    -0.07
    umor
    -0.06
     Love
    -0.06
    [j
    -0.06
    POSITIVE LOGITS
     section
    0.17
     sections
    0.16
    Section
    0.14
     Sections
    0.12
     Section
    0.12
    section
    0.10
     sectional
    0.10
    [section
    0.10
    Sections
    0.10
    _section
    0.09
    Act Density 0.022%

    No Known Activations