© 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. 20-GEMMASCOPE-RES-16K
    6. 15417
    Prev
    Next
    INDEX
    Explanations

    programming terms and function definitions related to tree data structures

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

    GUI and tree views

    np_acts-logits-general · gemini-2.0-flash

    GUI elements and controls

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    expiresIn
    -0.53
     conduire
    -0.49
    tvguidetime
    -0.48
     graag
    -0.47
     relever
    -0.46
     BitmapFactory
    -0.46
     createdAt
    -0.46
    componentWill
    -0.45
     tomber
    -0.45
     setEmail
    -0.45
    POSITIVE LOGITS
     GUI
    0.93
     UI
    0.90
     buttons
    0.85
     gui
    0.82
     dialog
    0.82
    GUI
    0.81
    gui
    0.81
    UI
    0.78
     menu
    0.77
     ui
    0.75
    Activations Density 0.622%

    No Known Activations