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. 40762
    Prev
    Next
    INDEX
    Explanations

    phrases emphasizing legal rulings and liability determinations

    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
     kaynağından
    -0.56
    ingtones
    -0.52
    oughby
    -0.49
    帖最后由
    -0.49
    XMLSchema
    -0.48
    setVerticalGroup
    -0.48
     veto
    -0.47
    uitton
    -0.47
    าสตร์
    -0.47
    UnusedPrivate
    -0.47
    POSITIVE LOGITS
     autorytatywna
    0.53
    TypedDataSet
    0.46
    ViewImports
    0.45
    RegressionTest
    0.43
    SharedCtor
    0.42
    IUrlHelper
    0.41
    ChildScrollView
    0.41
    ViewFeatures
    0.38
    :✨
    0.37
    DispatchToProps
    0.35
    Activations Density 0.003%

    No Known Activations