© 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. 11194
    Prev
    Next
    INDEX
    Explanations

    graphical representations and statistical analysis elements

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

    descriptions of geometric and plot-related elements in technical text, emphasizing shapes, coordinates, and spatial relationships in figures.

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

    words related to geometric and spatial concepts, particularly in mathematical or technical contexts describing lines, curves, points, circles, and spatial relationships.

    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
    ftagPool
    -0.38
     réforme
    -0.37
    vably
    -0.36
     AspNetCore
    -0.35
    UniformLocation
    -0.34
    djangoproject
    -0.34
    AlterField
    -0.33
     Variance
    -0.33
     Mekong
    -0.33
    anmoins
    -0.33
    POSITIVE LOGITS
    AndEndTag
    0.55
    adaptiveStyles
    0.49
    évaluateur
    0.49
    ंदीखरीदारी
    0.49
    hasErrors
    0.47
     graph
    0.47
     ویکی‌پدی
    0.46
    تقاوى
    0.45
    Enllaços
    0.45
     незавершена
    0.44
    Activations Density 0.657%

    No Known Activations