© 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. 119041
    Prev
    Next
    INDEX
    Explanations

    code syntax elements

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

    whitespace and line breaks in code, particularly after semicolons and opening braces in programming languages.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    This neuron detects the boilerplate initialization/disposal lines generated by Windows Forms designer (e.g. the `components = null;` assignment and `InitializeComponent()` call).

    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
     or
    -1.98
     to
    -1.66
     only
    -1.57
    only
    -1.49
     “
    -1.49
    になりますが
    -1.48
     then
    -1.48
    because
    -1.44
     Would
    -1.38
    Because
    -1.36
    POSITIVE LOGITS
    𝒰
    1.64
    klju
    1.56
     églises
    1.54
    換え
    1.52
     نیز
    1.52
    <bos>
    1.51
    varande
    1.47
    畕
    1.45
    
    1.45
    aI
    1.44
    Activations Density 0.003%

    No Known Activations