Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 5509
    Prev
    Next
    INDEX
    Explanations

    attends to the effect of parentheses from various mathematical computations and function tests

    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
    0:0.10
    1:0.12
    2:0.12
    3:0.13
    4:0.14
    5:0.11
    6:0.09
    7:0.15
    Negative Logits
     dAtA
    -0.33
     Millar
    -0.28
    NameInMap
    -0.25
    Iain
    -0.25
     canst
    -0.24
    ll
    -0.24
     Iain
    -0.23
    BuildContext
    -0.22
    WidgetItem
    -0.22
    śred
    -0.22
    POSITIVE LOGITS
    findpost
    0.45
    richTextPanel
    0.32
    ագրություններ
    0.32
    randomUUID
    0.31
     otomatig
    0.30
     Padma
    0.30
     חיצוניים
    0.30
    invokeLater
    0.28
    SourceChecksum
    0.28
    borderSide
    0.27
    Activations Density 0.178%

    No Known Activations