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-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 92194
    Prev
    Next
    INDEX
    Explanations

    structural and syntactical elements in legal documents

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

    we and Defendant

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    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
    Aiheesta
    -0.63
     kasarigan
    -0.59
    IntoConstraints
    -0.57
    principalColumn
    -0.55
    httphttps
    -0.54
    QMetaType
    -0.52
     IndexPath
    -0.50
    Lähteet
    -0.50
    PhysRev
    -0.49
     forState
    -0.49
    POSITIVE LOGITS
     Италијани
    0.36
     rum
    0.33
     vacacionales
    0.32
     chẳng
    0.31
    adpleegd
    0.29
    HtmlAttribute
    0.29
    bö
    0.29
     kaarangay
    0.29
    omla
    0.29
    eneuve
    0.28
    Activations Density 0.199%

    No Known Activations