© 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. 128236
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on numeric tokens—i.e. standalone numbers or digit sequences in code or text.

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

    code structures and foreign languages

    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
    parque
    -1.02
    度に
    -1.02
     ketika
    -0.95
    ⼥
    -0.92
    ubakar
    -0.92
     cometer
    -0.92
    timb
    -0.91
     utama
    -0.91
     vī
    -0.91
    político
    -0.90
    POSITIVE LOGITS
     sodio
    0.97
    最初
    0.94
     trots
    0.93
    社
    0.93
     lazily
    0.91
     +
    0.89
    begins
    0.89
    省略
    0.88
     هیچ
    0.85
     Taran
    0.85
    Activations Density 0.069%

    No Known Activations