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. 68842
    Prev
    Next
    INDEX
    Explanations

    terms related to legal proceedings and entitlements

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

    co, women, di, their, relevant, been, similarly, act, close, relative

    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
    Segoe
    -0.34
     Ex
    -0.32
    bootstrapcdn
    -0.31
    ForRow
    -0.31
     contextLoads
    -0.30
    Soo
    -0.30
     EX
    -0.29
    HasAnnotation
    -0.29
    Expl
    -0.29
    Demografie
    -0.28
    POSITIVE LOGITS
     <<<<<<<<<<<<<<
    0.54
    SharedDtor
    0.54
    AndEndTag
    0.49
     MainAxisSize
    0.49
     AssemblyCulture
    0.49
    PhysRev
    0.48
     дописавши
    0.47
    IBILITIES
    0.46
    //
    0.46
    RenderAtEndOf
    0.46
    Activations Density 0.198%

    No Known Activations