© 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. 68027
    Prev
    Next
    INDEX
    Explanations

    numbers and year citations

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

    The neuron fires on numeric tokens—especially years, dates, issue or volume numbers, and other reference identifiers in citations.

    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
    lâ
    -1.02
    här
    -1.02
     sostuvo
    -0.96
     ajustado
    -0.94
     confiar
    -0.94
     actions
    -0.93
     mög
    -0.91
    anmar
    -0.91
     мы
    -0.91
    appelle
    -0.91
    POSITIVE LOGITS
     condam
    1.00
    荭
    0.90
    nisses
    0.89
     menatap
    0.89
     marea
    0.88
     conexion
    0.88
     santuario
    0.88
     arka
    0.87
     недорого
    0.87
    Хотя
    0.85
    Activations Density 0.008%

    No Known Activations