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

    Neuronpedia

    APICircuit 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. 6324
    Prev
    Next
    INDEX
    Explanations

    attends to numerical or mathematical tokens marked with asterisks from tokens later represented by square brackets

    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
    1:0.09
    2:0.07
    3:0.07
    4:0.49
    5:0.07
    6:0.05
    7:0.06
    Negative Logits
    )";
    
    -0.42
     doInBackground
    -0.37
    Personendaten
    -0.36
    AccessorTable
    -0.35
    DispatchToProps
    -0.35
     betweenstory
    -0.34
    ]));
    
    -0.34
    ㎠
    -0.34
     **/
    
    -0.33
    ")));
    
    -0.33
    POSITIVE LOGITS
    ...
    0.33
    &
    0.27
    _
    0.26
    ....
    0.24
    '
    0.23
    </
    0.23
    +
    0.21
    --
    0.21
    esi
    0.21
    jo
    0.21
    Activations Density 0.134%

    No Known Activations