© 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. 104566
    Prev
    Next
    INDEX
    Explanations

    non-alphanumeric characters

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

    ASCII art and decorative text characters in source code comments and documentation.

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

    The neuron activates on lines of text composed of non‐alphanumeric symbols arranged in repeating patterns, i.e. ASCII art or decorative symbol‐based diagrams.

    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
     to
    -2.27
     after
    -2.11
     if
    -2.00
     for
    -1.89
    ↵↵↵
    -1.87
     may
    -1.81
     should
    -1.78
     ensures
    -1.77
     take
    -1.71
     it
    -1.66
    POSITIVE LOGITS
     Dinas
    1.88
     SqlDbType
    1.87
    actéristi
    1.85
    ᵴ
    1.71
    羮
    1.71
    
    1.66
    ʆ
    1.66
    mäh
    1.64
    mostly
    1.62
     EXPECT
    1.60
    Activations Density 0.008%

    No Known Activations