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. 6105
    Prev
    Next
    INDEX
    Explanations

    attends to variables or tokens used in mathematical expressions from preceding or related tokens in the context of equations

    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
    0:0.36
    1:0.18
    2:0.12
    3:0.05
    4:0.07
    5:0.07
    6:0.04
    7:0.07
    Negative Logits
     gomma
    -0.30
     européens
    -0.29
     européennes
    -0.28
     fondi
    -0.27
    his
    -0.27
     sanitarios
    -0.27
     ordini
    -0.26
     decorazione
    -0.26
     miei
    -0.25
     africains
    -0.25
    POSITIVE LOGITS
    principalColumn
    0.35
    Datuak
    0.32
    llong
    0.32
     continúas
    0.31
     OBITUARY
    0.31
    itates
    0.31
     linkovi
    0.31
    uksessa
    0.30
    ndorf
    0.30
    __(/*!
    0.30
    Activations Density 1.236%

    No Known Activations