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

    Neuronpedia

    INDEX
    Explanations

    topics or phrases related to programming and web development

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

    phrases that indicate the start of a new prompt or question, often identifying conversational openers or introductory context in both English and Chinese.

    oai_token-act-pair · gpt-4.1-2025-04-14Triggered by @ywhp2c8tjm

    the beginning of new documents or questions, especially those mentioning safety, security, or belonging to groups.

    oai_token-act-pair · gemini-2.0-flashTriggered by @ywhp2c8tjm

    Preceding "is," "may," or articles

    np_acts-logits-general · gemini-2.0-flash

    which followed by verbs

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

    conversational openings and direct questions—often about identity/definitions or expressing worry about crime victimization—across both English and Chinese.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
    ConstraintMaker
    -0.75
     ब्रेकडाउन
    -0.74
    KommentareTeilen
    -0.73
    MLLoader
    -0.70
    tagHelperRunner
    -0.70
    ScopeManager
    -0.70
     CreateTagHelper
    -0.69
    httphttps
    -0.69
     estekak
    -0.68
    IsContent
    -0.68
    POSITIVE LOGITS
    <bos>
    0.65
    The
    0.42
    <
    0.38
    A
    0.37
    <strong>
    0.37
    I
    0.35
    [
    0.35
    1
    0.34
    ↵↵
    0.34
    <h1>
    0.33
    Act Density 56.916%

    No Known Activations