Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 6661
    Prev
    Next
    INDEX
    Explanations

    attends to punctuation tokens from various preceding tokens in the structure of sentences

    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
    0:0.05
    1:0.08
    2:0.04
    3:0.04
    4:0.19
    5:0.51
    6:0.02
    7:0.03
    Negative Logits
    انيف
    -0.41
    fjspx
    -0.41
     typelib
    -0.41
     Paglinawan
    -0.40
    kwds
    -0.40
    tanleria
    -0.39
     חיצוניים
    -0.39
     تكبرها
    -0.38
    edicated
    -0.37
     <=",
    -0.36
    POSITIVE LOGITS
    OnPage
    0.25
     welcher
    0.23
     either
    0.22
    さえ
    0.22
    ControllerAdvice
    0.22
     both
    0.21
    ტერ
    0.21
    nad
    0.20
    ствий
    0.20
    both
    0.20
    Activations Density 1.338%

    No Known Activations