© 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. 22-GEMMASCOPE-RES-131K
    4. 84347
    Prev
    Next
    INDEX
    Explanations

    code delimiters and structure

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

    This neuron detects bracket symbols in code—both square brackets (e.g. array indexing) and angle brackets (e.g. templates).

    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
     because
    -1.36
     from
    -1.26
     just
    -1.03
     but
    -1.02
     campeón
    -1.00
    氾
    -0.98
    嚥
    -0.95
    蠱
    -0.95
     alimentation
    -0.95
    👛
    -0.95
    POSITIVE LOGITS
     salvaje
    1.13
    _
    
    1.08
    *}\
    1.07
     jeżeli
    1.05
    [])
    
    0.97
    '){
    
    0.97
    /}
    0.97
     verdaderas
    0.96
    ")));
    0.96
    >());
    0.93
    Activations Density 0.013%

    No Known Activations