© 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. 7503
    Prev
    Next
    INDEX
    Explanations

    numbers, ratios, quantities, and timeframes, especially having to do with medical studies or events

    oai_token-act-pair · gemini-2.0-flash

    archaic or old-fashioned English words

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

    older scientific texts

    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
     ever
    -0.91
     et
    -0.66
     Ever
    -0.65
     –
    -0.65
     [
    -0.57
     EVER
    -0.56
     u
    -0.55
     v
    -0.55
     A
    -0.54
     I
    -0.54
    POSITIVE LOGITS
     myſelf
    1.21
     ſeveral
    1.18
     itſelf
    1.17
     Monfieur
    1.05
     Efq
    1.03
     auffi
    1.02
     pleaſure
    1.01
     whoſe
    1.00
     ſche
    1.00
     iſt
    0.96
    Activations Density 2.246%

    No Known Activations