© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 4-GEMMASCOPE-TRANSCODER-16K
    7. 9423
    Prev
    Next
    INDEX
    Explanations

    comments in code that indicate work to be done

    oai_token-act-pair · gemini-2.0-flash

    TODO

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_4/width_16k/average_l0_88
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.4.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    <bos>
    -0.71
    الإنجليزية
    -0.70
    IntoConstraints
    -0.70
    parsedMessage
    -0.60
    yarnpkg
    -0.59
     onOptions
    -0.57
     وتسجيلات
    -0.56
     resourceCulture
    -0.55
     AssemblyCulture
    -0.54
     organisation
    -0.52
    POSITIVE LOGITS
    ="@+
    0.67
     Wikimedijinoj
    0.63
    laut
    0.56
    escalier
    0.54
     Савезне
    0.54
    SEGUIR
    0.54
     commandés
    0.54
     Teg
    0.52
     TextAlign
    0.51
    RepeatedField
    0.51
    Activations Density 3.528%

    No Known Activations