© 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. 22-GEMMASCOPE-RES-131K
    7. 128700
    Prev
    Next
    INDEX
    Explanations

    The neuron detects isolated numeric tokens, particularly large standalone numbers (e.g. figure or table reference numbers).

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

    citation markers

    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
    //
    -1.05
     its
    -1.00
    想必
    -0.95
     zuletzt
    -0.95
    ету
    -0.93
    getYear
    -0.91
     school
    -0.89
     ざ
    -0.88
    確認ください
    -0.88
    SPECTION
    -0.87
    POSITIVE LOGITS
    obatan
    1.20
    があります
    1.16
    もあります
    1.09
    
    1.08
    らしいです
    1.07
    viewDid
    1.06
     DbType
    1.00
    fitriones
    1.00
     réinitial
    0.99
    Kiri
    0.98
    Activations Density 0.005%

    No Known Activations