© 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. 39227
    Prev
    Next
    INDEX
    Explanations

    holy titles and orders

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

    The neuron is specialized for recognizing formal religious or liturgical terminology (e.g. names of sacred texts, hymns, prayers, and honorifics like “holy” or “blessed”).

    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
     jengibre
    -1.35
    UREAU
    -1.25
    ńst
    -1.24
     gewel
    -1.23
    
    -1.20
     ochran
    -1.18
    不明白
    -1.18
    áneo
    -1.18
     включая
    -1.16
    🡺
    -1.16
    POSITIVE LOGITS
    They
    1.50
     they
    1.30
    Another
    1.20
     зробити
    1.17
     time
    1.16
    Despite
    1.13
    Histor
    1.12
    Even
    1.10
    Almost
    1.09
    Already
    1.07
    Activations Density 0.054%

    No Known Activations