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. 6391
    Prev
    Next
    INDEX
    Explanations

    attends to tokens representing a specific scientific or technical entity from corresponding tokens that provide further classification or specification

    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
    0:0.10
    1:0.37
    2:0.09
    3:0.10
    4:0.09
    5:0.07
    6:0.06
    7:0.09
    Negative Logits
    InjectAttribute
    -0.29
     majeur
    -0.23
    isEnd
    -0.20
    resizingMask
    -0.20
     архивлан
    -0.20
    Hentet
    -0.20
    Y
    -0.20
     taches
    -0.20
     end
    -0.19
     zależności
    -0.19
    POSITIVE LOGITS
    #+#
    0.52
    IsMutable
    0.43
    httphttps
    0.39
     незавершена
    0.37
     ExecuteAsync
    0.36
     referenties
    0.36
     мәкал
    0.35
    oa̍t
    0.34
     Audiodateien
    0.34
    iastes
    0.34
    Activations Density 0.583%

    No Known Activations