© 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. 65550
    Prev
    Next
    INDEX
    Explanations

    html tags and attributes

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

    Explanation of neuron 4 behavior: the main thing this neuron does is find HTML markup elements (tags and attribute tokens).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     meticulous
    -1.47
     in
    -1.45
     then
    -1.38
     it
    -1.29
     by
    -1.27
     meticulously
    -1.27
     specific
    -1.23
    -1.23
     adequate
    -1.21
     with
    -1.20
    POSITIVE LOGITS
     figliu
    1.65
     rengi
    1.52
     cedo
    1.47
     fenô
    1.46
     Insgesamt
    1.45
     fechar
    1.43
    ſo
    1.39
     leitor
    1.39
     abraço
    1.38
     trazer
    1.34
    Activations Density 0.179%
    No Known Activations