© 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. 34-GEMMASCOPE-RES-131K
    7. 106559
    Prev
    Next
    INDEX
    Explanations

    add, alter table

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

    This neuron strongly activates on numeric tokens (digit sequences) in code (e.g. version numbers, SQL numeric literals, migration IDs).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    疃
    -0.81
    ########.
    -0.73
    Osi
    -0.73
    므
    -0.71
    Initializing
    -0.69
    HELLO
    -0.69
    bov
    -0.68
     glow
    -0.68
    Submitting
    -0.68
    entered
    -0.67
    POSITIVE LOGITS
     adding
    1.56
     add
    1.50
    ALTER
    1.45
     alteration
    1.44
     altering
    1.43
    添加
    1.34
     alter
    1.34
     alterations
    1.31
     ALTER
    1.30
    alter
    1.29
    Activations Density 0.011%

    No Known Activations