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. 6039
    Prev
    Next
    INDEX
    Explanations

    attends to the token "Section" from tokens describing various sections in a structured document

    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
    0:0.05
    1:0.53
    2:0.04
    3:0.02
    4:0.06
    5:0.16
    6:0.04
    7:0.03
    Negative Logits
     section
    -0.43
     Section
    -0.42
    Section
    -0.33
     “
    -0.31
     Chapter
    -0.27
     unit
    -0.26
     Page
    -0.26
     
    -0.25
     Part
    -0.25
    {}",
    -0.25
    POSITIVE LOGITS
     disambiguazione
    0.58
    IsContent
    0.51
    rungsseite
    0.50
    ########.
    0.50
     متعلقه
    0.46
     InputDecoration
    0.46
    SourceChecksum
    0.44
     autorytatywna
    0.42
    +#+#
    0.42
    UrlResolution
    0.41
    Activations Density 0.054%

    No Known Activations