© 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. 11914
    Prev
    Next
    INDEX
    Explanations

    the word "early" and terms associated with it such as "childhood", and "late"

    oai_token-act-pair · gemini-2.0-flash

    terms related to childhood

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

    early

    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
    <?
    -0.60
    kháu
    -0.60
     myſelf
    -0.59
     кӀ
    -0.59
     vectoriales
    -0.59
    UserScript
    -0.58
     équ
    -0.57
    IContainer
    -0.56
     comprends
    -0.55
     tịch
    -0.55
    POSITIVE LOGITS
    WebServlet
    0.54
    tilli
    0.52
     access
    0.51
    __((
    0.49
     childhood
    0.48
     last
    0.47
     preso
    0.45
    quirer
    0.45
    ConstraintMaker
    0.45
    iliation
    0.45
    Activations Density 1.271%

    No Known Activations