© 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. 34722
    Prev
    Next
    INDEX
    Explanations

    numbered or listed items

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

    The neuron strongly activates on numeric and punctuation tokens that form volume, page, year, and section markers—i.e. the numeric parts of legal or academic 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_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
    inya
    -0.77
    XD
    -0.76
    Ито
    -0.73
    onStop
    -0.73
    ː
    -0.72
     judgement
    -0.71
     staining
    -0.71
     enää
    -0.70
    — 
    -0.70
    rowa
    -0.69
    POSITIVE LOGITS
     Part
    0.95
    Part
    0.92
    ("[
    0.85
     INTRODUCTION
    0.85
    Professor
    0.85
    INTRODUCTION
    0.84
     Professor
    0.84
     Assistant
    0.81
    ckel
    0.76
     “[
    0.76
    Activations Density 0.017%

    No Known Activations