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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 9-GEMMASCOPE-RES-16K
    6. 9485
    Prev
    Next
    INDEX
    Explanations

    geographic names, particularly those related to U.S. states and cities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 9-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.9.hook_resid_post
    Hook Layer
    9
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    herself
    -0.62
    ազմ
    -0.61
     Sorg
    -0.60
    ⚭
    -0.58
    tschaft
    -0.57
     ostavi
    -0.57
    +#+#
    -0.57
    ManyToOne
    -0.56
    cestry
    -0.56
    ANNES
    -0.56
    POSITIVE LOGITS
     Pennsylvania
    1.05
     Florida
    1.03
     Wisconsin
    1.01
     Georgia
    1.00
     Colorado
    1.00
     Texas
    0.98
     Maryland
    0.97
    Florida
    0.95
     Illinois
    0.94
     Massachusetts
    0.93
    Activations Density 0.211%

    No Known Activations