© 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. 54493
    Prev
    Next
    INDEX
    Explanations

    table, model, Q, {, [, Yemen

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

    The neuron fires on the literal “table” keyword (often in DDL statements like CREATE TABLE), i.e. occurrences of database table declarations.

    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
     calça
    -1.05
     Cuando
    -1.02
     délé
    -1.01
     garrafa
    -1.00
     fidélité
    -0.99
    Cuando
    -0.99
     toalha
    -0.98
    mappedBy
    -0.98
     vôtre
    -0.97
     toalla
    -0.96
    POSITIVE LOGITS
     is
    1.21
    ⅙
    1.19
     in
    1.18
     after
    1.10
     on
    1.09
    ยน
    1.07
     different
    1.05
    ⅕
    1.05
     at
    1.02
     >>=
    1.00
    Activations Density 0.000%

    No Known Activations