© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 9526
    Prev
    Next
    INDEX
    Explanations

    class definitions

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

    The neuron fires strongly on Java code declaration tokens—especially modifiers (like “volatile”) and class or constructor identifier names—i.e. the beginnings of field, class, or constructor definitions.

    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
    蟥
    -1.34
    lampen
    -1.34
    Kons
    -1.32
    stande
    -1.28
     dekoratif
    -1.26
    junge
    -1.25
    鬟
    -1.25
    
    -1.23
    醅
    -1.23
     唯美
    -1.20
    POSITIVE LOGITS
    );
    1.44
    ();
    1.42
     '
    1.38
    ">
    1.14
    }
    1.14
    XMLSchema
    1.12
    T
    1.12
    /*
    1.12
    le
    1.11
    1.07
    Activations Density 0.008%

    No Known Activations