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. 5850
    Prev
    Next
    INDEX
    Explanations

    attends to various structural or syntactical elements in programming or markup languages, particularly focusing on tokens like brackets and specific keywords

    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
    1:0.12
    2:0.24
    3:0.16
    4:0.10
    5:0.05
    6:0.05
    7:0.14
    Negative Logits
    ⓧ
    -0.53
     Paglinawan
    -0.42
    Życiorys
    -0.41
    oredCriteria
    -0.40
    ".
    
    -0.40
    出版年
    -0.39
    SharedCtor
    -0.38
    ).
    
    -0.38
     jScrollPane
    -0.38
    ngilizce
    -0.37
    POSITIVE LOGITS
    "
    0.28
     خارجية
    0.27
    //
    0.24
     Table
    0.24
    Toro
    0.23
    theon
    0.23
    שו
    0.22
    Table
    0.22
     "
    0.22
    '
    0.22
    Activations Density 0.000%

    No Known Activations