© 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. 128182
    Prev
    Next
    INDEX
    Explanations

    The neuron is looking for programming‐oriented tokens (e.g. code keywords, file extensions, script/include directives) that indicate the presence of source code or scripts.

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

    numbers and dates

    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
     Subsid
    -0.89
    )->
    -0.85
    IZABETH
    -0.84
     leſs
    -0.84
    🧂
    -0.80
    隍
    -0.78
     hro
    -0.77
    āja
    -0.76
    )});
    -0.76
     masih
    -0.75
    POSITIVE LOGITS
     artikkelen
    0.88
     elétrica
    0.86
    tache
    0.85
     its
    0.83
     mė
    0.78
    peluche
    0.78
     sogenannte
    0.76
    叹了口气
    0.76
    usen
    0.75
    olished
    0.75
    Activations Density 0.213%

    No Known Activations