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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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 - 16k
    5. 10-GEMMASCOPE-RES-16K
    6. 4566
    Prev
    Next
    INDEX
    Explanations

    specific aims or objectives in research proposals

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    îtra
    -0.61
    WireFormatLite
    -0.58
     Ngb
    -0.57
    دانشنامهٔ
    -0.56
    aarrggbb
    -0.55
     <",
    -0.55
    ArrowToggle
    -0.54
    /-/
    -0.53
     ElementRef
    -0.52
     florales
    -0.52
    POSITIVE LOGITS
    
    0.64
    RenderAtEndOf
    0.61
     unknownFields
    0.60
     Lastly
    0.59
     conclusion
    0.59
    zelfde
    0.59
     }}"></
    0.58
     Also
    0.58
    وفي
    0.58
     lastly
    0.56
    Activations Density 0.490%

    No Known Activations