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. 5907
    Prev
    Next
    INDEX
    Explanations

    attends to references within single elements from token elements referred to in the context of their respective databases or groups

    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
    0:0.18
    1:0.23
    2:0.19
    3:0.04
    4:0.03
    5:0.03
    6:0.03
    7:0.22
    Negative Logits
     سكانية
    -0.42
    存于互联网档案馆
    -0.40
     дописавши
    -0.37
     חיצוניים
    -0.37
    #+#
    -0.35
    
    -0.34
    /*
    -0.31
    ("")]
    
    -0.31
    __(/*!
    -0.31
    /**
    -0.30
    POSITIVE LOGITS
    LabelTagHelper
    0.25
    foreground
    0.22
    AdapterView
    0.22
     Woodrow
    0.22
    zeba
    0.22
     Sharjah
    0.21
     solidari
    0.21
     cifra
    0.20
    XtraReports
    0.20
     nämlich
    0.20
    Activations Density 0.196%

    No Known Activations