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. 5962
    Prev
    Next
    INDEX
    Explanations

    attends to deficiency-related tokens from paired tokens related to other patterns

    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
    0:0.14
    1:0.16
    2:0.11
    3:0.10
    4:0.09
    5:0.04
    6:0.15
    7:0.16
    Negative Logits
     noDo
    -0.31
     @"/
    -0.29
     Inscrivez
    -0.26
    aung
    -0.26
     feltro
    -0.25
     miel
    -0.25
    chieht
    -0.24
    sstil
    -0.24
    ubung
    -0.24
     AIO
    -0.24
    POSITIVE LOGITS
    зулта
    0.41
     '\\;'
    0.39
    yntaxException
    0.38
    MigrationBuilder
    0.35
    AndEndTag
    0.32
    yarnpkg
    0.31
    EDEFAULT
    0.30
    RegistryLite
    0.30
    kháu
    0.30
    ViewFeatures
    0.29
    Activations Density 0.036%

    No Known Activations