© 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. 129804
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on SQL statement keywords, especially those used in UPDATE clauses (e.g. “SET” and “WHERE”).

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

    SQL WHERE clause

    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
    fokus
    -1.86
     retten
    -1.66
     enorme
    -1.60
     ideen
    -1.58
     decidi
    -1.56
     skolen
    -1.55
     aprendizagem
    -1.54
     mété
    -1.53
     mortes
    -1.52
     elegante
    -1.52
    POSITIVE LOGITS
     in
    1.77
    </em>
    1.70
     at
    1.59
    </b>
    1.41
     electrica
    1.36
     with
    1.34
     chemic
    1.33
     within
    1.28
     without
    1.25
     not
    1.23
    Activations Density 0.006%

    No Known Activations