© 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. 102014
    Prev
    Next
    INDEX
    Explanations

    prologue or epilogue

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

    The neuron fires on section or chapter headings in texts—words like Prologue, Prelude, Interlude, and Epilogue.

    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
    盦
    -2.11
     you
    -2.00
    
    -1.98
    ፕ
    -1.85
    柺
    -1.85
    zehnten
    -1.82
    跸
    -1.82
     cambiando
    -1.82
    最新的
    -1.81
     sorprend
    -1.79
    POSITIVE LOGITS
    嫆
    2.31
    If
    2.08
    itsubishi
    2.08
    The
    2.06
    This
    2.02
    ruguay
    2.02
    淪
    2.00
    ꦮ
    2.00
    鼕
    1.95
    部分は
    1.94
    Activations Density 0.008%

    No Known Activations