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. 6169
    Prev
    Next
    INDEX
    Explanations

    attends to verbs indicating actions related to ownership or transactions from subjects that reference these actions

    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
    0:0.05
    1:0.11
    2:0.07
    3:0.07
    4:0.19
    5:0.32
    6:0.07
    7:0.07
    Negative Logits
    elski
    -0.25
    UserScript
    -0.24
     OMITTED
    -0.23
    eterangan
    -0.23
    repr
    -0.23
     resourceCulture
    -0.22
     gynhyrchwyd
    -0.22
     putea
    -0.21
    antastic
    -0.21
     Type
    -0.21
    POSITIVE LOGITS
    Jeografia
    0.40
    CopyWith
    0.38
     вікі
    0.36
    EndProject
    0.35
     MaterialApp
    0.35
    شنایی
    0.34
    openConnection
    0.33
     requestCode
    0.33
    JMenu
    0.33
    elemField
    0.32
    Activations Density 0.063%

    No Known Activations