© 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. 130570
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on standalone numeric tokens in business/legal text—e.g. street numbers, suite numbers, zip-codes, and other numeric identifiers.

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

    city or state abbreviations

    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
    žené
    -0.92
     أيضا
    -0.89
    ajes
    -0.85
    ₑ
    -0.85
     tombol
    -0.84
    ḿ
    -0.83
     Benth
    -0.82
     firſt
    -0.82
     kaarangay
    -0.81
    lln
    -0.81
    POSITIVE LOGITS
    DU
    0.97
     Kitchen
    0.93
     altri
    0.93
    Дереккөздер
    0.89
     Dube
    0.85
     NULL
    0.85
     Greece
    0.85
    رامی
    0.85
     DU
    0.85
     تصریح
    0.84
    Activations Density 0.016%

    No Known Activations