© 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. 11146
    Prev
    Next
    INDEX
    Explanations

    code snippets and special characters

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

    code syntax elements, particularly punctuation and special characters used in programming languages like dollar signs, brackets, slashes, and function/variable declarations.

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

    This neuron fires on programming-style tokens and syntax (e.g. code punctuation like $, /*, #, backslashes, XML tags, keywords like int/for, etc.), i.e. it detects bits of source code.

    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
     all
    -1.64
     to
    -1.52
    强大的
    -1.25
     as
    -1.25
    Enzo
    -1.22
     conseguido
    -1.22
    Thanos
    -1.19
     refugio
    -1.16
    Verte
    -1.15
     SALUD
    -1.14
    POSITIVE LOGITS
     ';
    
    1.32
     TextBox
    1.17
    utica
    1.17
    "},
    
    1.16
    WO
    1.16
    陽
    1.13
     již
    1.13
    ITS
    1.13
    Poznám
    1.13
    TextBox
    1.12
    Activations Density 0.167%

    No Known Activations