© 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. 3015
    Prev
    Next
    INDEX
    Explanations

    words or phrases related to numeric values, measurements, finance, regulations, and quantities of physical resources

    oai_token-act-pair · gemini-2.0-flash

    programming keywords and syntax

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

    technical writing

    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
    then
    -0.62
     then
    -0.54
    Then
    -0.53
     infrequently
    -0.51
     normally
    -0.51
    THEN
    -0.49
     tehn
    -0.48
     då
    -0.47
     Cardinale
    -0.47
    նակ
    -0.47
    POSITIVE LOGITS
     is
    0.94
    expandindo
    0.82
     AssemblyCulture
    0.75
     Chwiliwch
    0.71
     becomes
    0.69
    oredCriteria
    0.68
    InitStruct
    0.67
    rungsseite
    0.64
     Becomes
    0.61
     '\\;'
    0.61
    Activations Density 17.695%

    No Known Activations