© 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. 59238
    Prev
    Next
    INDEX
    Explanations

    WPF/UI controls and programming

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

    The neuron is picking out programming identifiers—especially capitalized class, control, and type names (e.g. ListBox, Canvas, ItemsPanel, ViewModel)—in code snippets.

    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
    どの
    -1.30
     gourmand
    -1.11
     같은
    -1.09
     fasti
    -1.07
     geene
    -1.06
     sabem
    -1.06
     bian
    -1.01
     oogen
    -0.96
    같
    -0.96
     from
    -0.95
    POSITIVE LOGITS
     still
    1.38
     this
    1.33
     would
    1.27
     yine
    1.26
     utama
    1.25
    葑
    1.23
    ἵ
    1.23
     will
    1.21
     autobuses
    1.20
     प्र
    1.16
    Activations Density 0.184%

    No Known Activations