© 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. 21-GEMMASCOPE-TRANSCODER-16K
    7. 15521
    Prev
    Next
    INDEX
    Explanations

    code block endings in various programming languages, particularly the closing curly brace

    oai_token-act-pair · gemini-2.0-flash

    code block closures

    np_token-act-pair-logits · gpt-4o-mini

    }

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    principalColumn
    -0.92
    tagHelperRunner
    -0.77
    httphttps
    -0.73
    Diweddarwch
    -0.71
    '])){
    -0.71
    )";
    -0.68
    DoubleQuotes
    -0.68
    '}),
    -0.67
    EndInit
    -0.65
    出版年
    -0.65
    POSITIVE LOGITS
     Italij
    0.49
    Cau
    0.45
     elettrica
    0.43
    用户信息
    0.42
    cud
    0.41
     więcej
    0.41
     électro
    0.41
     elett
    0.40
    칸
    0.40
    acter
    0.40
    Activations Density 0.395%

    No Known Activations