© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 10368
    Prev
    Next
    INDEX
    Explanations

    terms related to scientific and academic concepts, particularly in the context of research and education

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    category labels and references sections in structured documents.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    Wikipedia-style metadata and glossary lines—especially category tags and acronym/abbreviation definitions with colons or semicolons.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    structured metadata and reference sections, particularly categories, abbreviations, and bibliographic information.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     initComponents
    -0.52
    قایناق‌لار
    -0.50
    rungsseite
    -0.38
    roupa
    -0.35
    waffe
    -0.33
     competition
    -0.33
    lapsingToolbar
    -0.33
    Външни
    -0.33
    Rujukan
    -0.33
    Vem
    -0.33
    POSITIVE LOGITS
    PerformLayout
    0.68
     kaarangay
    0.68
    󠁴
    0.66
     Infórmanos
    0.62
     Савезне
    0.60
    最快更新
    0.57
     otomatig
    0.57
    ьаж
    0.56
     ivelany
    0.54
    новниш
    0.52
    Activations Density 0.126%

    No Known Activations