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 - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 16338
    Prev
    Next
    INDEX
    Explanations

    programming language constructs and structures related to event handling and initialization of modules

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     propOrder
    -0.75
     AssemblyCompany
    -0.71
    principalColumn
    -0.67
    :✨
    -0.62
     nakalista
    -0.60
    ConstraintMaker
    -0.57
    SizeMode
    -0.56
     ostavi
    -0.54
    delwed
    -0.54
     Normdatei
    -0.53
    POSITIVE LOGITS
     spontaneous
    0.42
     Olympic
    0.40
    chige
    0.40
    exports
    0.39
     spontaneously
    0.39
     ResponseEntity
    0.38
     métro
    0.36
     Erben
    0.36
    <tfoot>
    0.36
    endmodule
    0.35
    Activations Density 0.099%

    No Known Activations