Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 21-GEMMASCOPE-RES-16K
    6. 11546
    Prev
    Next
    INDEX
    Explanations

    numerical data and formatting in code

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.hook_resid_post
    Hook Layer
    21
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    лтамалар
    -0.41
    GEBURTSDATUM
    -0.39
    rikan
    -0.33
    vestream
    -0.33
     MainAxisSize
    -0.32
    iffion
    -0.32
    alamus
    -0.32
    мирование
    -0.31
     Brag
    -0.31
     TRABAJO
    -0.31
    POSITIVE LOGITS
    CppMethod
    0.56
    queryInterface
    0.54
     okuyayım
    0.54
    今度は
    0.49
     StatelessWidget
    0.48
    TEP
    0.48
    enderror
    0.47
    again
    0.46
     szint
    0.46
    DoubleQuotes
    0.45
    Activations Density 0.055%

    No Known Activations