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. 10-GEMMASCOPE-ATT-16K
    6. 6074
    Prev
    Next
    INDEX
    Explanations

    attends to tokens containing "navbar" and "method" from programming-related 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
    0:0.09
    1:0.08
    2:0.05
    3:0.06
    4:0.33
    5:0.24
    6:0.05
    7:0.05
    Negative Logits
    owieck
    -0.29
     koş
    -0.23
     vincit
    -0.23
    kuper
    -0.23
    messageInfo
    -0.23
     Estatal
    -0.23
    \
    -0.22
    arms
    -0.22
     olímp
    -0.22
    Viitteet
    -0.22
    POSITIVE LOGITS
     createSlice
    0.46
    tagHelperRunner
    0.45
    basicConfig
    0.43
    Portale
    0.43
     intptr
    0.41
    LookAnd
    0.40
    inflater
    0.39
     EconPapers
    0.39
    ніципалі
    0.38
    }{*}{}
    0.37
    Activations Density 0.001%

    No Known Activations