© 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. 65559
    Prev
    Next
    INDEX
    Explanations

    code comments and declarations

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

    This neuron specifically detects programming-code tokens such as language identifiers, annotations/macros, class or attribute names, version numbers, and other non-natural-language symbols.

    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
     from
    -1.55
     other
    -1.41
     many
    -1.26
     some
    -1.23
     several
    -1.19
     in
    -1.18
     nine
    -1.12
     their
    -1.10
     most
    -1.09
     four
    -1.09
    POSITIVE LOGITS
    美味しいです
    1.23
     schuhe
    1.20
    about
    1.20
    呦
    1.18
     difficoltà
    1.16
     判断
    1.16
     apenas
    1.14
     فقط
    1.13
    vantaged
    1.13
     TAMBÉM
    1.09
    Activations Density 0.001%

    No Known Activations