© 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. 1358
    Prev
    Next
    INDEX
    Explanations

    Android activity and builder patterns

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

    This neuron detects Android/Java API identifiers and method‐call tokens (e.g. class names like Builder, Fragment, Activity, LayoutParams and methods like beginTransaction, addToBackStack, replace).

    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
     erhi
    -1.75
    凩
    -1.64
    使用了
    -1.54
    OWE
    -1.52
    .".
    -1.50
    ."'
    -1.49
    …(
    -1.49
    *;
    -1.48
     ventures
    -1.48
     garantía
    -1.48
    POSITIVE LOGITS
    Nearly
    1.63
    Almost
    1.61
    Основные
    1.55
    Several
    1.50
    Taking
    1.47
    There
    1.46
    Although
    1.46
    Причины
    1.43
    Shortly
    1.41
    Before
    1.41
    Activations Density 0.003%

    No Known Activations