© 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. 10875
    Prev
    Next
    INDEX
    Explanations

    Public facing data structures

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

    The neuron selectively fires on code‐style identifiers or constants—e.g. long CamelCase names, all‐caps names (often with underscores), and similar embedded code tokens.

    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
     francia
    -0.96
     dán
    -0.88
     limpeza
    -0.86
     peruana
    -0.83
     pierwszej
    -0.83
    ventajas
    -0.82
     filtros
    -0.80
     skupina
    -0.80
     február
    -0.80
     advierte
    -0.78
    POSITIVE LOGITS
     spars
    0.94
     triliun
    0.87
    +)$
    0.85
     carriera
    0.84
     dina
    0.84
    🔭
    0.83
    Assignable
    0.82
     価格
    0.82
    wydd
    0.82
    mathbb
    0.82
    Activations Density 0.001%

    No Known Activations