© 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. 7677
    Prev
    Next
    INDEX
    Explanations

    assign a

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

    The neuron selectively activates on numeric tokens (digits, years, measurements, etc.), effectively spotting numbers in the text.

    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
    ๑
    0.57
     রি
    0.55
    kyl
    0.55
     ハ
    0.54
     Ма
    0.53
     カラ
    0.53
     guarding
    0.53
     Фу
    0.51
     ム
    0.51
    agnetic
    0.50
    POSITIVE LOGITS
    UTILS
    0.64
    helloworld
    0.60
    EXIST
    0.59
    眇
    0.59
    喩
    0.59
     बखूबी
    0.58
    偌
    0.58
    全世界
    0.57
    有一个
    0.57
    気づ
    0.57
    Activations Density 0.397%

    No Known Activations