© 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. 38585
    Prev
    Next
    INDEX
    Explanations

    code directives and punctuation

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

    The neuron strongly activates on punctuation tokens—especially periods (including ellipses) and colons—marking sentence‐ or section‐boundary markers.

    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
     示例
    -1.00
     может
    -0.98
    versaire
    -0.90
     galle
    -0.90
    mmons
    -0.88
     Bahan
    -0.87
    ＄
    -0.87
    actéristiques
    -0.85
    just
    -0.84
    iendo
    -0.84
    POSITIVE LOGITS
     ざ
    1.20
     thậm
    1.05
     emphas
    1.01
     usw
    0.99
     andern
    0.98
    demás
    0.94
    において
    0.92
    atized
    0.90
     vz
    0.89
    dars
    0.89
    Activations Density 0.014%

    No Known Activations