© 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. 9607
    Prev
    Next
    INDEX
    Explanations

    code closing braces and keywords

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

    This neuron fires on lines that close out C/C++ preprocessor conditional blocks (e.g. “#endif”).

    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
     dudes
    -0.97
     magasin
    -0.94
    جغرافيا
    -0.89
    ÁND
    -0.88
    liyor
    -0.88
     poème
    -0.88
    Mvh
    -0.86
    bedaan
    -0.86
     bonheur
    -0.85
    
    -0.85
    POSITIVE LOGITS
     все
    0.94
    l
    0.94
    la
    0.91
    a
    0.88
    场所
    0.85
    ла
    0.84
    cepteur
    0.84
    !
    0.84
    A
    0.84
    ums
    0.82
    Activations Density 0.010%

    No Known Activations