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-2-2B
    4. Residual Stream - 65k
    5. 17-GEMMASCOPE-RES-65K
    6. 28561
    Prev
    Next
    INDEX
    Explanations

    phrases describing adult behavior and experiences, particularly in contexts of injury or medical conditions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.17.hook_resid_post
    Hook Layer
    17
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    MeasureSpec
    -0.52
    TestMethod
    -0.51
    flexDirection
    -0.47
    Datuak
    -0.45
    getRuntime
    -0.45
    SizeF
    -0.45
    Portail
    -0.45
    ResponseWriter
    -0.44
     these
    -0.43
     обо
    -0.42
    POSITIVE LOGITS
     ddelweddau
    0.77
    Hentet
    0.75
    ItemBackground
    0.71
    AsUp
    0.68
     unui
    0.64
     seorang
    0.64
    uage
    0.60
    KindOfClass
    0.59
     kogo
    0.59
    PreferredItem
    0.59
    Activations Density 0.134%

    No Known Activations