© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 36-GEMMASCOPE-RES-16K
    11. 6154
    Prev
    Next
    INDEX
    Explanations

    references to cities and their capitals

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.36.hook_resid_post
    Hook Layer
    36
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Walkover
    -0.47
    AndEndTag
    -0.44
    ContentAlignment
    -0.43
     MainAxisSize
    -0.42
     Wikimedijinoj
    -0.42
    Clik
    -0.41
    complexContent
    -0.40
     aDecoder
    -0.40
    WebServlet
    -0.39
    ChromeDriver
    -0.39
    POSITIVE LOGITS
     capital
    3.30
    capital
    2.91
     Capital
    2.86
    Capital
    2.77
     CAPITAL
    2.75
    CAPITAL
    2.50
     capitol
    2.39
     capitale
    2.38
     capitals
    2.34
     cap
    2.33
    Activations Density 0.613%

    No Known Activations