© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 70901
    Prev
    Next
    INDEX
    Explanations

    numbers formatted with commas

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

    The neuron is specialized in detecting numeric quantities (e.g. numbers, figures, counts, dates) in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    ↵
    -2.95
    to
    -1.91
     most
    -1.85
     When
    -1.83
     has
    -1.71
     During
    -1.70
    section
    -1.67
     sehr
    -1.66
     Give
    -1.65
    -
    -1.62
    POSITIVE LOGITS
     escritores
    1.83
    僦
    1.80
     "",
    1.80
     .......
    1.76
     músicos
    1.74
    >();
    
    1.71
    printLine
    1.69
    ())),
    1.69
    :"
    1.67
     první
    1.66
    Activations Density 0.025%

    No Known Activations