© 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. 28641
    Prev
    Next
    INDEX
    Explanations

    mathematical and code delimiters

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

    The neuron is selectively activating on tokens that are parts of mathematical notation—numbers, variables, operators, and punctuation used in equations.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    Đây
    -0.92
     приходи
    -0.91
     sogget
    -0.90
     this
    -0.90
     zones
    -0.85
     impieg
    -0.85
    lju
    -0.85
     movimenti
    -0.84
     教程
    -0.84
     пиро
    -0.84
    POSITIVE LOGITS
    тій
    0.99
     BIBLIOGRAPHY
    0.96
    〕
    0.91
    ?>">
    0.90
    ).[
    0.87
    клон
    0.86
     CWE
    0.85
    ']):
    0.85
    properly
    0.85
    PUESTA
    0.84
    Activations Density 0.039%

    No Known Activations