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. 6052
    Prev
    Next
    INDEX
    Explanations

    attends to dishes containing specific ingredients marked as notable from ingredient mentions earlier in the sequence

    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
    1:0.10
    2:0.09
    3:0.13
    4:0.09
    5:0.03
    6:0.19
    7:0.23
    Negative Logits
     déchir
    -0.31
     reaſon
    -0.31
     reafon
    -0.29
     Theſe
    -0.29
     ainfi
    -0.29
     supérieurs
    -0.29
     pattes
    -0.28
    jbl
    -0.27
     εφ
    -0.27
     catég
    -0.27
    POSITIVE LOGITS
    новниш
    0.34
     doInBackground
    0.32
    исленность
    0.31
     HttpNotFound
    0.28
    CodedInputStream
    0.28
    AddTagHelper
    0.28
    kháu
    0.27
    AsUp
    0.27
     Italijanski
    0.27
    [...,
    0.27
    Activations Density 0.029%

    No Known Activations