© 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. 10-GEMMASCOPE-RES-131K
    7. 105770
    Prev
    Next
    INDEX
    Explanations

    section markers and titles

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

    The neuron detects isolated numeric tokens (section numbers, label IDs, reference markers, etc.) embedded in markup commands.

    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
    ߳
    -1.68
     essenciais
    -1.66
    鳎
    -1.65
    もありました
    -1.65
     verlangen
    -1.56
    vailability
    -1.54
    してました
    -1.45
     libération
    -1.45
    Comes
    -1.42
    ܆
    -1.42
    POSITIVE LOGITS
    the
    2.30
    as
    2.05
    
    2.02
    un
    1.95
     sebagai
    1.92
    ala
    1.86
    ệu
    1.84
    all
    1.78
     mereka
    1.73
    ATE
    1.72
    Activations Density 0.001%

    No Known Activations