© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 38342
    Prev
    Next
    INDEX
    Explanations

    programming language statements

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

    The neuron spikes on programming‐language type or class names (e.g. Java imports, C++ includes, Scala futures)—in other words, it detects identifiers of classes, interfaces, or types in 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_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
    ゙
    -0.88
     most
    -0.88
    izarse
    -0.87
     diverse
    -0.81
     is
    -0.81
    ять
    -0.76
     resolute
    -0.76
     umumnya
    -0.76
     tender
    -0.73
     invalid
    -0.73
    POSITIVE LOGITS
    PhysRevD
    1.07
    hasMoreElements
    0.99
     conſ
    0.90
     فبراير
    0.88
    enseignement
    0.86
    ⃝
    0.86
    Ᏺ
    0.86
    °.
    0.86
     dépass
    0.85
    ?".
    0.84
    Activations Density 0.013%

    No Known Activations