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. 4
    Prev
    Next
    INDEX
    Explanations

    attends to the token "stage" from tokens labeled as different stages

    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
    0:0.70
    1:0.01
    2:0.01
    3:0.03
    4:0.15
    5:0.03
    6:0.01
    7:0.03
    Negative Logits
     
    -0.40
     of
    -0.33
    ntö
    -0.25
     (
    -0.24
     La
    -0.24
     so
    -0.23
    labus
    -0.23
     la
    -0.22
     l
    -0.22
     ...
    -0.21
    POSITIVE LOGITS
     CreateTagHelper
    0.59
    findpost
    0.52
     surla
    0.52
    makeConstraints
    0.50
    yntaxException
    0.50
    fillType
    0.47
     }}"></
    0.46
     ―――――
    0.46
     myſelf
    0.44
     initComponents
    0.44
    Activations Density 0.016%

    No Known Activations