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. 5903
    Prev
    Next
    INDEX
    Explanations

    attends to numerical values from special character tokens in the given sequence

    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
    1:0.10
    2:0.11
    3:0.11
    4:0.10
    5:0.09
    6:0.18
    7:0.18
    Negative Logits
    ListItemIcon
    -0.32
    Composable
    -0.32
    feira
    -0.30
    boj
    -0.30
    anță
    -0.29
    chossen
    -0.28
     Rennen
    -0.28
     Roper
    -0.28
     culturelles
    -0.28
     Maier
    -0.27
    POSITIVE LOGITS
    RegressionTest
    0.39
    nastics
    0.31
    })]
    0.30
    (!__
    0.30
    cydow
    0.30
    prepareStatement
    0.29
    ulose
    0.28
    LoggerFactory
    0.28
    Datuak
    0.28
     $("<
    0.28
    Activations Density 0.007%

    No Known Activations