© 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-2-2B
    5. 21-GEMMASCOPE-TRANSCODER-16K
    7. 6463
    Prev
    Next
    INDEX
    Explanations

    words that make writing more formal and bureaucratic

    oai_token-act-pair · gemini-2.0-flash

    mentions of government officials or bodies

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

    US government

    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
    Not in Any Lists

    No Comments

    Negative Logits
    spesies
    -0.56
    kloped
    -0.53
    BeginInit
    -0.51
     Moors
    -0.49
     Kirke
    -0.48
    ListItemIcon
    -0.47
     anzuzeigen
    -0.47
    ttemberg
    -0.47
    execSQL
    -0.47
    namentales
    -0.47
    POSITIVE LOGITS
     Congress
    0.95
     President
    0.88
     congress
    0.87
    President
    0.80
    congress
    0.80
    Congress
    0.79
    addCriterion
    0.74
     president
    0.73
    Cong
    0.70
     Secretary
    0.68
    Activations Density 7.721%

    No Known Activations