© 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. 11866
    Prev
    Next
    INDEX
    Explanations

    code snippets related to graph algorithms and structure traversal

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
    不做
    -0.38
    俱
    -0.36
     LEYENDO
    -0.36
     occafion
    -0.35
     distanciation
    -0.34
    LabelTagHelper
    -0.34
    érience
    -0.34
    İŞ
    -0.34
    gues
    -0.33
    newpage
    -0.33
    POSITIVE LOGITS
     betweenstory
    0.59
    GEBURTSDATUM
    0.57
    ValueGenerated
    0.52
     ModelRenderer
    0.51
    endmodule
    0.49
    hoeddwyd
    0.47
    makeConstraints
    0.45
    mappedBy
    0.45
     AppComponent
    0.44
     évêque
    0.44
    Activations Density 0.242%

    No Known Activations