© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129586
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on numeric tokens—i.e. it detects numbers (sequences of digits, including integers or floats) in the text.

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

    low values or quantities

    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
    guous
    -0.85
    zków
    -0.79
     destul
    -0.77
    zanie
    -0.75
    比特
    -0.75
    MFC
    -0.73
     ovate
    -0.71
     urgencia
    -0.69
    KPI
    -0.69
    uzzy
    -0.69
    POSITIVE LOGITS
     low
    1.12
     fewer
    1.09
    خفض
    1.00
     reduced
    0.98
     downgraded
    0.96
     shrunk
    0.94
     downgrade
    0.92
     reductions
    0.91
    shrink
    0.90
    reduced
    0.88
    Activations Density 0.114%

    No Known Activations