© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 54274
    Prev
    Next
    INDEX
    Explanations

    code blocks or delimiters

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

    The neuron flags programming “header” tokens that begin new sections or blocks—e.g. Read, begin/BEGIN, public, document—i.e. code‐block or statement starters.

    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
    providedIn
    -1.11
     schlafen
    -1.09
    føring
    -1.02
    🕺
    -1.01
     Besoin
    -0.98
    関する
    -0.97
    有り
    -0.96
     full
    -0.96
     Compare
    -0.95
    ;
    -0.94
    POSITIVE LOGITS
    で
    1.11
    では
    1.05
    臵
    0.97
    (
    0.96
     hefty
    0.94
    でも
    0.93
     TextAlign
    0.93
    ۹
    0.91
     eatery
    0.90
    は
    0.90
    Activations Density 0.002%

    No Known Activations