© 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. 21-GEMMASCOPE-TRANSCODER-16K
    7. 2629
    Prev
    Next
    INDEX
    Explanations

    mixed-language (especially French and English) internet chat

    oai_token-act-pair · gemini-2.0-flash

    varied language or code terms

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

    Technical discussions, mixed languages

    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
    vably
    -0.59
    ьогодні
    -0.56
    adə
    -0.52
    rungsseite
    -0.51
     obstante
    -0.50
     թվական
    -0.50
    together
    -0.49
    kuuta
    -0.49
    StackNavigator
    -0.48
    endgroup
    -0.47
    POSITIVE LOGITS
    contentLoaded
    0.79
    AndroidJUnit
    0.69
    rrggbb
    0.60
    さ
    0.59
     shawl
    0.59
    NameInMap
    0.57
    "}")
    0.57
    θα
    0.56
    Diweddarwch
    0.56
     Galile
    0.56
    Activations Density 1.143%

    No Known Activations