Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 30615
    Prev
    Next
    INDEX
    Explanations

    terminology related to balls and their motion

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

    bouncing ball

    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
    SourceChecksum
    -0.65
    ftagPool
    -0.63
     ujednoznacz
    -0.63
     invokingState
    -0.60
    AndEndTag
    -0.57
     lenker
    -0.55
    MemoryWarning
    -0.52
     wikipagina
    -0.52
    IsMutable
    -0.50
     ainfi
    -0.50
    POSITIVE LOGITS
     ball
    0.41
     hit
    0.34
    ball
    0.34
     Sing
    0.33
    IsVisible
    0.33
     balls
    0.33
    큼
    0.32
    交
    0.31
     Ball
    0.30
    欠
    0.30
    Activations Density 0.003%

    No Known Activations