© 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. 8987
    Prev
    Next
    INDEX
    Explanations

    references to figures and tables in the document

    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
    <unused68>
    -0.72
    <unused52>
    -0.72
    <unused16>
    -0.72
    <unused8>
    -0.72
    <unused17>
    -0.71
    <unused23>
    -0.71
    <unused41>
    -0.71
    <unused42>
    -0.71
    <pad>
    -0.71
    [@BOS@]
    -0.71
    POSITIVE LOGITS
    Fig
    0.52
    Figure
    0.50
    setVerticalGroup
    0.45
    KommentareTeilen
    0.40
    Рис
    0.37
     Fig
    0.37
    FIG
    0.35
     Relaciones
    0.33
     Penga
    0.33
    évaluateur
    0.32
    Activations Density 0.360%

    No Known Activations