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. 5713
    Prev
    Next
    INDEX
    Explanations

    attends to the conjunction "so" from various contextual clauses that elaborate or explain preceding statements

    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
    0:0.12
    1:0.45
    2:0.10
    3:0.05
    4:0.06
    5:0.05
    6:0.04
    7:0.08
    Negative Logits
     further
    -0.27
    ofern
    -0.27
     so
    -0.25
     فع
    -0.25
    Revenir
    -0.25
     snart
    -0.24
    Để
    -0.24
     adelante
    -0.24
    lyk
    -0.23
    бяза
    -0.23
    POSITIVE LOGITS
    expandindo
    0.56
    webElementXpaths
    0.56
    Autoritní
    0.48
    numerusform
    0.48
    HasAnnotation
    0.47
    EDEFAULT
    0.45
    TypedDataSet
    0.44
     MessageBoxIcon
    0.43
    AsUp
    0.43
    CodeAttribute
    0.42
    Activations Density 0.386%

    No Known Activations