© 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. 130318
    Prev
    Next
    INDEX
    Explanations

    The neuron fires strongly on web‐ and HTML‐related tokens—things like “html,” “http,” “web,” “browser,” and other markup or protocol keywords.

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

    web and HTML

    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
    pegno
    -1.06
    IDATE
    -0.90
    pejo
    -0.86
    hoenix
    -0.86
     levels
    -0.86
    したので
    -0.85
    antula
    -0.84
    参数
    -0.82
    pogon
    -0.82
    ovina
    -0.82
    POSITIVE LOGITS
     web
    1.81
     website
    1.48
    Www
    1.39
     webpage
    1.38
     HTML
    1.38
     online
    1.38
     WEB
    1.34
     html
    1.32
     Web
    1.31
    Html
    1.29
    Activations Density 0.261%

    No Known Activations