© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 14651
    Prev
    Next
    INDEX
    Explanations

    references and code comments

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

    The neuron detects code‐ or markup‐style identifiers and attribute names (e.g. camelCase or snake_case tokens, XML/HTML attributes, and other programming identifiers).

    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
     if
    -2.44
     when
    -2.34
     because
    -2.05
     before
    -1.98
    {
    -1.95
     regardless
    -1.88
     there
    -1.86
     providing
    -1.83
     often
    -1.82
     after
    -1.77
    POSITIVE LOGITS
    尨
    1.78
     senza
    1.74
     également
    1.73
     trám
    1.72
    </sup>
    1.70
     overarching
    1.70
    
    1.69
     "";
    1.68
     WORLD
    1.68
     Algunos
    1.67
    Activations Density 0.000%

    No Known Activations