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. 6137
    Prev
    Next
    INDEX
    Explanations

    attends to the token "junk" from related tokens linked to the context of sport and entertainment

    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
    0:0.04
    1:0.06
    2:0.05
    3:0.08
    4:0.07
    5:0.03
    6:0.57
    7:0.07
    Negative Logits
     ๆ
    -0.32
     &
    -0.32
    βολ
    -0.31
     contener
    -0.30
    ítja
    -0.30
    ită
    -0.30
    ítása
    -0.30
    pture
    -0.29
     na
    -0.29
    duit
    -0.28
    POSITIVE LOGITS
    Datuak
    0.59
    InputBorder
    0.58
     незавершена
    0.55
     referrerpolicy
    0.54
    UrlResolution
    0.54
     doInBackground
    0.54
     оригіналу
    0.52
     سكانية
    0.51
     насељу
    0.49
    rungsseite
    0.49
    Activations Density 0.181%

    No Known Activations