© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 0-GEMMASCOPE-TRANSCODER-16K
    7. 9490
    Prev
    Next
    INDEX
    Explanations

    a mix of Russian words related to web development and words related to code attributes

    oai_token-act-pair · gemini-2.0-flash

    Russian or Java/C++ code

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_0/width_16k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.0.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     rhod
    -0.84
     armp
    -0.73
     morire
    -0.73
     Myco
    -0.71
     néglig
    -0.71
     myco
    -0.70
     foncé
    -0.70
     lyre
    -0.69
     perox
    -0.68
     Tuba
    -0.67
    POSITIVE LOGITS
     the
    0.89
     estekak
    0.81
    точно
    0.60
    таратура
    0.59
    чь
    0.57
    UnsafeEnabled
    0.57
    точ
    0.56
     some
    0.56
     by
    0.55
    щи
    0.55
    Activations Density 5.321%

    No Known Activations