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. 5712
    Prev
    Next
    INDEX
    Explanations

    attends to dependency-related tokens from tokens following the term "dependency."

    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
    0:0.15
    1:0.34
    2:0.09
    3:0.07
    4:0.09
    5:0.08
    6:0.04
    7:0.10
    Negative Logits
     nakalista
    -0.53
    IonicModule
    -0.50
    AddTagHelper
    -0.49
    MessageOf
    -0.49
    awtextra
    -0.48
    ValueStyle
    -0.47
    >{@
    -0.45
    GEBURTSDATUM
    -0.44
    addPreferredGap
    -0.44
    principalTable
    -0.44
    POSITIVE LOGITS
    ly
    0.42
     tombé
    0.36
     sở
    0.33
    OrNil
    0.33
    G
    0.32
     intégr
    0.31
     chers
    0.31
     WebDriverWait
    0.30
    όρ
    0.30
     juicio
    0.30
    Activations Density 0.254%

    No Known Activations