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-2-2B
    4. Attention Out - 16k
    5. 10-GEMMASCOPE-ATT-16K
    6. 6388
    Prev
    Next
    INDEX
    Explanations

    attends to date-related tokens from other numerical tokens

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 10-gemmascope-att-16k
    Configuration
    google/gemma-scope-2b-pt-att/layer_10/width_16k/average_l0_70
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.10.attn.hook_z
    Hook Layer
    10
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.28
    1:0.21
    2:0.20
    3:0.04
    4:0.06
    5:0.10
    6:0.03
    7:0.05
    Negative Logits
     Galen
    -0.39
    elry
    -0.32
    Selección
    -0.32
    &__
    -0.31
     unknownFields
    -0.31
     Eggs
    -0.31
     orel
    -0.30
    CAREER
    -0.29
    liesslich
    -0.29
     kedua
    -0.29
    POSITIVE LOGITS
    ViewFeatures
    0.50
    migrationBuilder
    0.41
    astify
    0.38
     AssemblyTitle
    0.38
     onCreateView
    0.38
    ynnwys
    0.37
     DPRD
    0.36
    oluted
    0.35
    complexContent
    0.34
    windowFixed
    0.34
    Activations Density 0.082%

    No Known Activations