© 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-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 113155
    Prev
    Next
    INDEX
    Explanations

    references to updates and messages in a technical or programming context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    Telegram updates

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Stuhl
    -0.38
     cockpit
    -0.36
     mái
    -0.36
    횟
    -0.35
    getWriter
    -0.35
     Lak
    -0.35
    IBIL
    -0.35
     Jes
    -0.35
     Arne
    -0.35
    Wy
    -0.34
    POSITIVE LOGITS
     Telegram
    0.75
    telegram
    0.72
    Telegram
    0.71
     telegram
    0.65
    #
    0.54
    NameInMap
    0.54
    HasAnnotation
    0.52
    migrationBuilder
    0.50
    Spoljašnje
    0.50
     DevExpress
    0.47
    Activations Density 0.002%

    No Known Activations