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. 5783
    Prev
    Next
    INDEX
    Explanations

    attends to comma-related tokens marked at the end of sequences from the comma-related tokens present in previous positions

    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
    0:0.06
    1:0.19
    2:0.07
    3:0.03
    4:0.23
    5:0.27
    6:0.04
    7:0.06
    Negative Logits
    migrationBuilder
    -0.44
    principalColumn
    -0.41
    #
    -0.33
     resourceCulture
    -0.33
     MediatR
    -0.32
    تقاوى
    -0.31
    NSIndexPath
    -0.29
    queryInterface
    -0.28
    __*/
    -0.27
    ')}}"
    -0.27
    POSITIVE LOGITS
    helves
    0.24
     Referential
    0.24
    URIComponent
    0.23
    hysema
    0.23
    dresser
    0.23
     Ababa
    0.23
     Butterflies
    0.22
    saida
    0.22
    csol
    0.22
    ראי
    0.21
    Activations Density 0.072%

    No Known Activations