© 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. 84092
    Prev
    Next
    INDEX
    Explanations

    holy and sacred things

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

    The neuron fires on honorific or sacred terms and titles (e.g. “Holy,” “Sacred,” “Enthroned,” “Virgin,” “Gracious,” etc.) indicating religious reverence.

    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
    inher
    -0.86
     ancaman
    -0.84
    Erstellt
    -0.84
    狴
    -0.84
    licante
    -0.82
    ТО
    -0.82
     werkt
    -0.82
     goûts
    -0.82
     アクリル
    -0.82
     utilisateurs
    -0.82
    POSITIVE LOGITS
     one
    1.27
     ones
    1.08
     One
    1.03
    One
    0.90
     keeping
    0.89
     ресни
    0.87
     actions
    0.86
    lète
    0.85
     Kriege
    0.85
     Substanz
    0.84
    Activations Density 0.015%

    No Known Activations