© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 1574
    Prev
    Next
    INDEX
    Explanations

    HTML structure and web elements

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

    The neuron is specially tuned to spot HTML‐style code tokens—HTML tags (e.g. `<a>`), attribute names (e.g. `id`, `class`), quotation marks, and other markup syntax.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Tief
    0.67
     whanne
    0.63
     Grupo
    0.58
     Distillery
    0.58
     Reaktion
    0.55
     shouldUse
    0.55
     Canonical
    0.54
    ᴸ
    0.54
     Wur
    0.54
     Morel
    0.54
    POSITIVE LOGITS
    mysqli
    0.98
     mysqli
    0.95
    navbar
    0.93
    Navbar
    0.92
     <?
    0.84
    topnav
    0.84
    <?
    0.83
    marquee
    0.82
    stylesheet
    0.80
    <html>
    0.79
    Activations Density 0.003%

    No Known Activations