© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128779
    Prev
    Next
    INDEX
    Explanations

    The neuron detects HTML/XML markup, i.e. the angle‐bracketed tag openings (the “<” and tag/attribute tokens).

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

    code and user interface elements

    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
    orship
    -0.91
     other
    -0.89
    ॰
    -0.89
    ighed
    -0.88
     kazak
    -0.87
     zara
    -0.86
     sond
    -0.86
     Romagna
    -0.85
     лв
    -0.85
    ਿਆ
    -0.84
    POSITIVE LOGITS
     brinde
    0.91
    цуз
    0.88
     föränd
    0.87
    斧
    0.85
     föruts
    0.85
     cuarentena
    0.84
     tidigare
    0.82
     competitively
    0.82
    ᷄
    0.81
     millennial
    0.81
    Activations Density 0.004%

    No Known Activations