© 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 - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 8857
    Prev
    Next
    INDEX
    Explanations

    CSS properties and values

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

    Color terms or "user" token

    np_acts-logits-general · gemini-2.0-flash

    red color

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 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
     initComponents
    -0.81
     مرئيه
    -0.64
     ModelExpression
    -0.64
     JpaRepository
    -0.63
    ########.
    -0.62
    ImageContext
    -0.61
    AnchorStyles
    -0.61
    RegressionTest
    -0.59
    IntoConstraints
    -0.59
     otomatig
    -0.59
    POSITIVE LOGITS
     color
    0.40
    Blue
    0.40
     Blue
    0.40
    Color
    0.39
     Color
    0.39
     Colors
    0.37
     Yellow
    0.36
     blue
    0.36
     yellow
    0.36
     colors
    0.35
    Activations Density 0.113%

    No Known Activations