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. 0-GEMMASCOPE-ATT-16K
    6. 211
    Prev
    Next
    INDEX
    Explanations

    attends to "target" tokens from programming or scientific contexts from related tokens also marked as "target."

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 0-gemmascope-att-16k
    Configuration
    google/gemma-scope-2b-pt-att/layer_0/width_16k/average_l0_104
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.0.attn.hook_z
    Hook Layer
    0
    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
    0:0.04
    1:0.75
    2:0.04
    3:0.01
    4:0.04
    5:0.04
    6:0.02
    7:0.02
    Negative Logits
    InputBorder
    -0.26
     Ibid
    -0.26
     tege
    -0.26
    حياته
    -0.24
     Pante
    -0.24
    HtmlAttribute
    -0.23
    TagHelper
    -0.23
    BindingResult
    -0.23
     XNUMX
    -0.23
    Ibid
    -0.23
    POSITIVE LOGITS
    ########.
    0.37
    లాలు
    0.37
    AddTagHelper
    0.31
    AndroidJUnit
    0.29
    Personensuche
    0.29
    chengladbach
    0.28
    VertexAttrib
    0.27
     betweenstory
    0.27
    těte
    0.27
    adaptiveStyles
    0.27
    Activations Density 2.754%

    No Known Activations