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

    Neuronpedia

    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    This neuron identifies section headings and their markers (e.g., “INTRODUCTION,” “DISCUSSION,” “Sec 1-1,” etc.) in academic-style documents.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     BS
    -0.07
     zero
    -0.07
     tp
    -0.06
     supermarket
    -0.06
    gel
    -0.06
    hendis
    -0.06
    रण
    -0.06
     usuario
    -0.06
    .googleapis
    -0.06
     biodiversity
    -0.06
    POSITIVE LOGITS
    <Member
    0.06
    ABL
    0.06
     encoded
    0.06
    Washington
    0.06
     pizzas
    0.05
     topical
    0.05
    .ResponseEntity
    0.05
    ^K
    0.05
     Pornhub
    0.05
    ιν
    0.05
    Act Density 0.003%

    No Known Activations