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. 5819
    Prev
    Next
    INDEX
    Explanations

    attends to closing square brackets that follow specific numeric markers from numbers and symbols that appear in the text

    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
    0:0.09
    1:0.15
    2:0.10
    3:0.12
    4:0.13
    5:0.12
    6:0.10
    7:0.15
    Negative Logits
    AndEndTag
    -0.39
    __*/
    -0.39
    TagMode
    -0.35
     ostavi
    -0.31
    COUVER
    -0.30
    Билгалдахарш
    -0.30
    AdapterFactory
    -0.30
    行
    -0.29
    chu
    -0.29
     Thad
    -0.29
    POSITIVE LOGITS
     defaultstate
    0.26
    ibald
    0.26
     BEDROOM
    0.25
    彡
    0.25
     Petite
    0.25
     WILSON
    0.24
     Beres
    0.24
     pneumoniae
    0.23
    パンマン
    0.23
     Undeclared
    0.23
    Activations Density 0.035%

    No Known Activations