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

    Neuronpedia

    INDEX
    Explanations

    formatting for lists and emphasis

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

    Markdown formatting markers such as bullets, bold/emphasis, headings, and similar structural tokens.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     codons
    0.16
     explosives
    0.16
     antigens
    0.15
     analogs
    0.15
     isotopes
    0.15
     quanta
    0.15
     ligands
    0.15
     smoothies
    0.15
     goblin
    0.15
     dragons
    0.15
    POSITIVE LOGITS
    <h3>
    0.19
    <blockquote>
    0.18
    ar
    0.18
    ap
    0.17
    az
    0.17
    ust
    0.17
    ada
    0.17
    ak
    0.17
    ul
    0.16
    This
    0.16
    Act Density 3.313%

    No Known Activations