© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 118751
    Prev
    Next
    INDEX
    Explanations

    code blocks and structure

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

    closing curly braces in code, particularly in JSX/JavaScript object literals and template syntax.

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

    This neuron specializes in spotting code-related punctuation tokens—especially curly braces ( `{` / `}` ) and ellipses (`...`) used in programming templates and expressions.

    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
      
    -1.93
     either
    -1.86
     it
    -1.64
     any
    -1.56
     is
    -1.53
     as
    -1.52
     regard
    -1.50
    -
    -1.42
     estiver
    -1.40
     Fußballspieler
    -1.37
    POSITIVE LOGITS
    2
    1.94
    3
    1.78
    7
    1.76
     ciebie
    1.74
    6
    1.70
    {
    
    1.67
    …」
    1.66
     höheren
    1.65
    8
    1.65
     urges
    1.63
    Activations Density 0.004%

    No Known Activations