© 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. 34-GEMMASCOPE-RES-131K
    7. 53897
    Prev
    Next
    INDEX
    Explanations

    Culver City

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

    The neuron responds to standalone numeric tokens—especially numbers in legal citations, dates, and document identifiers.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    阴阳
    -0.77
    áját
    -0.77
     ברי
    -0.75
    Hybrid
    -0.72
     tiles
    -0.71
    cester
    -0.70
    hybrid
    -0.68
     hybrid
    -0.67
     setEmail
    -0.66
    hlander
    -0.66
    POSITIVE LOGITS
     Boro
    0.69
    koon
    0.68
    reis
    0.67
     DOCUMENT
    0.66
     zakon
    0.65
    鹑
    0.65
     Są
    0.65
     Nguyễn
    0.64
     PAL
    0.64
    cM
    0.63
    Activations Density 0.058%

    No Known Activations