© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 16150
    Prev
    Next
    INDEX
    Explanations

    proper nouns and names related to various media and technology

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

    brand and product names like LEGO, Samsung GALAXY, Microsoft, and S.U. (Stampin' Up).

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi
    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
     voeten
    -0.36
    igreja
    -0.36
    ViewImports
    -0.35
     gevo
    -0.35
     independente
    -0.33
     gesprek
    -0.32
     ammat
    -0.32
    kon
    -0.32
    te
    -0.32
     spoloč
    -0.32
    POSITIVE LOGITS
     propOrder
    0.85
    ſelf
    0.63
    phosa
    0.63
    imageshack
    0.62
    vician
    0.60
     Tartar
    0.59
    aarrggbb
    0.59
    jago
    0.57
    ActionCreators
    0.56
    󠁴
    0.55
    Activations Density 0.434%

    No Known Activations