© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129739
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on HTML tag syntax—i.e. the angle‐bracket tokens (“<”, “>”, “</…>”) and tag names (like “textarea,” “div,” “span,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    HTML element tags and structure

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     praticamente
    -1.12
     same
    -1.06
     adatto
    -1.04
     facilmente
    -1.02
     chris
    -1.02
     eccellente
    -1.01
    accor
    -1.01
     verhind
    -1.00
     rapidamente
    -0.98
     certamente
    -0.97
    POSITIVE LOGITS
     подготовка
    0.97
     seductive
    0.92
    jestel
    0.92
     will
    0.89
    谐
    0.87
    ait
    0.84
     fanciful
    0.83
    hoz
    0.83
    並び
    0.82
    mektedir
    0.82
    Activations Density 0.017%

    No Known Activations