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. 6607
    Prev
    Next
    INDEX
    Explanations

    attends to the type-related tokens from arbitrary subsequent tokens

    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
    0:0.07
    1:0.21
    2:0.09
    3:0.06
    4:0.28
    5:0.08
    6:0.06
    7:0.10
    Negative Logits
     iconLine
    -0.26
     CreateTagHelper
    -0.26
    UniformLocation
    -0.25
     '\\;'
    -0.25
    leyebilirsiniz
    -0.24
     onCreateView
    -0.24
    viewDidLoad
    -0.23
    LogLevel
    -0.22
    สบ
    -0.22
    ñador
    -0.22
    POSITIVE LOGITS
    argout
    0.32
     dAtA
    0.32
     Grat
    0.31
    AddHtmlAttribute
    0.31
     precedent
    0.31
     NgModule
    0.30
     CommonModule
    0.30
     فريبيس
    0.30
     ddelweddau
    0.30
    المكان
    0.29
    Activations Density 0.332%

    No Known Activations