© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 58059
    Prev
    Next
    INDEX
    Explanations

    HTML tags and code keywords

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

    The neuron strongly activates on programming identifiers or symbols that start with a capital letter (e.g. class, type, or constant names in code).

    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
    Materiale
    -1.95
     frança
    -1.81
     siè
    -1.70
     parís
    -1.66
     persiste
    -1.65
     ALLA
    -1.64
     火鍋
    -1.63
     rocha
    -1.61
     indien
    -1.60
     transforme
    -1.60
    POSITIVE LOGITS
    _
    2.02
    L
    2.00
    M
    1.98
    H
    1.95
    In
    1.93
    F
    1.92
    A
    1.91
    J
    1.91
    R
    1.89
    T
    1.88
    Activations Density 0.030%

    No Known Activations