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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 68130
    Prev
    Next
    INDEX
    Explanations

    references to technical specifications and applications across various fields

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

    AV integration, hunting, physics, backpacks, lighting, home, businesses, radio, Defence, cars, surgery

    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
     seguido
    -0.40
    JScrollPane
    -0.37
    tigkeits
    -0.35
     initComponents
    -0.35
    rebro
    -0.35
    etzal
    -0.34
    positorio
    -0.33
     Tash
    -0.33
    genstein
    -0.33
    いただきます
    -0.32
    POSITIVE LOGITS
    ########.
    0.49
    \{\\
    0.48
     purpoſe
    0.45
    IsMutable
    0.44
    ServletException
    0.43
    protoimpl
    0.42
     pleaſure
    0.42
    存于互联网档案馆
    0.42
    presto
    0.41
     الحره
    0.41
    Activations Density 0.040%

    No Known Activations