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. 6244
    Prev
    Next
    INDEX
    Explanations

    attends to possessive pronouns from corresponding nouns

    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
    1:0.14
    2:0.42
    3:0.06
    4:0.04
    5:0.02
    6:0.03
    7:0.10
    Negative Logits
    mişti
    -0.31
     dimenti
    -0.30
    Hale
    -0.29
     gild
    -0.28
    bind
    -0.28
    Darren
    -0.27
     betra
    -0.27
     LHS
    -0.27
     golpea
    -0.27
    кора
    -0.27
    POSITIVE LOGITS
    mobileqq
    0.32
    ieteur
    0.28
    lihood
    0.26
    apimachinery
    0.26
    ">//
    0.26
     виправивши
    0.26
     EconPapers
    0.25
    moschino
    0.24
    Thrown
    0.24
    ladin
    0.24
    Activations Density 0.371%

    No Known Activations