© 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. 10975
    Prev
    Next
    INDEX
    Explanations

    variety of things

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

    This neuron fires on quantifying or superlative expressions—tokens that convey amounts, “wide variety,” “most X,” numbers, or other count/degree markers in marketing or descriptive text.

    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
     three
    -1.82
     two
    -1.73
     four
    -1.52
     eight
    -1.52
     seven
    -1.43
     we
    -1.42
     five
    -1.42
     nine
    -1.41
     fifteen
    -1.39
     fourteen
    -1.32
    POSITIVE LOGITS
     créer
    1.30
     définir
    1.09
     костюм
    1.08
     also
    1.06
     !!!!
    1.05
    糨
    1.05
     atrop
    1.04
     wilfully
    1.03
     rigue
    1.03
    ツヤ
    1.02
    Activations Density 0.002%

    No Known Activations