© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 11794
    Prev
    Next
    INDEX
    Explanations

    Swift code constructs

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

    The neuron fires on occurrences of programming‐language type names and data‐structure identifiers (e.g. Swift’s String, Array, Dictionary, NSString, NSArray, NSDictionary).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    쯜
    0.66
    тук
    0.65
     चौधरी
    0.64
     medir
    0.63
    <unused40>
    0.62
    స్తూ
    0.61
    មើ
    0.61
     calculer
    0.61
     중요하다
    0.61
     coder
    0.60
    POSITIVE LOGITS
     [[
    1.05
     [
    0.97
    [[
    0.94
    NSString
    0.92
    =[
    0.88
     NSString
    0.86
     [[[
    0.86
    dispatch
    0.82
    BOOL
    0.79
    :[
    0.79
    Activations Density 0.004%

    No Known Activations