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. 6100
    Prev
    Next
    INDEX
    Explanations

    attends to the token "to" from variations of the token "or" or similar conjunctions

    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
    0:0.47
    1:0.15
    2:0.10
    3:0.03
    4:0.04
    5:0.02
    6:0.03
    7:0.12
    Negative Logits
    świ
    -0.33
    Ӏ
    -0.33
     newBuilder
    -0.32
     چگونه
    -0.31
    うえ
    -0.30
     onResponse
    -0.30
     metropolitana
    -0.29
     bianchi
    -0.29
     attenuation
    -0.29
    inals
    -0.28
    POSITIVE LOGITS
    ніципалі
    0.58
    tagext
    0.52
    IsContent
    0.47
    \}\\
    0.47
    '));
    
    0.47
     "..\..\..\
    0.47
     CreateTagHelper
    0.45
    ьаж
    0.45
    }}$\\
    0.44
     initComponents
    0.44
    Activations Density 1.245%

    No Known Activations