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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 12595
    Prev
    Next
    INDEX
    Explanations

    specific terms and processes related to optimization and measurement in experimental contexts

    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
     CURIAM
    -0.49
    ⟬
    -0.47
    новниш
    -0.44
     obiet
    -0.43
    AsUp
    -0.43
    ResponseWriter
    -0.42
    gahan
    -0.40
     obiec
    -0.40
    ClientRect
    -0.39
     gebildet
    -0.38
    POSITIVE LOGITS
     search
    0.68
     Search
    0.62
     searches
    0.61
     searching
    0.60
    Search
    0.56
     SEARCH
    0.55
     búsqueda
    0.52
     Searching
    0.52
     搜索
    0.50
    search
    0.49
    Activations Density 1.009%

    No Known Activations