© 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. 27498
    Prev
    Next
    INDEX
    Explanations

    document structure or code

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

    The neuron selectively activates on standalone code or markup keywords and identifiers (like “document,” “SYSTEM,” “#endif,” “quit,” “bar,” etc.) that serve as structural markers in programming or typesetting snippets.

    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
    -1.33
     then
    -1.21
     睡
    -1.20
     lembaga
    -1.17
     also
    -1.13
     of
    -1.11
    витие
    -1.09
     differ
    -1.08
     among
    -1.05
     berbagai
    -1.05
    POSITIVE LOGITS
     prohi
    1.22
    !
    1.20
    Honestly
    1.19
    incontro
    1.13
    larınız
    1.10
    地點
    1.09
     hề
    1.09
    Going
    1.09
    Adding
    1.09
    chufe
    1.09
    Activations Density 0.001%

    No Known Activations