© 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. 22-GEMMASCOPE-RES-131K
    7. 28094
    Prev
    Next
    INDEX
    Explanations

    semicolon and code syntax

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

    The neuron detects occurrences of programming class or type names (i.e. CamelCase identifiers).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    ใหญ
    -0.95
     for
    -0.95
    the
    -0.92
     with
    -0.88
     Dr
    -0.85
     riport
    -0.82
    zig
    -0.82
    hata
    -0.82
    aurus
    -0.81
     along
    -0.79
    POSITIVE LOGITS
    grö
    0.93
    quien
    0.91
    Yuki
    0.91
    Tür
    0.88
    Kura
    0.85
    ㆁ
    0.85
     membrana
    0.85
    Usos
    0.84
    Sakura
    0.83
     diğer
    0.83
    Activations Density 0.003%

    No Known Activations