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. 0-GEMMASCOPE-ATT-16K
    6. 112
    Prev
    Next
    INDEX
    Explanations

    attends to tokens that are enclosed in brackets from tokens later in the sequence

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 0-gemmascope-att-16k
    Configuration
    google/gemma-scope-2b-pt-att/layer_0/width_16k/average_l0_104
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.0.attn.hook_z
    Hook Layer
    0
    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
    0:0.02
    1:0.06
    2:0.06
    3:0.10
    4:0.12
    5:0.06
    6:0.46
    7:0.07
    Negative Logits
     betweenstory
    -1.23
    ")));
    
    -1.19
    )");
    
    -1.16
    ']],
    -1.14
    }")
    
    -1.14
    "},
    
    -1.12
     tartalomajánló
    -1.11
    '},
    
    -1.11
     beginnetje
    -1.10
    "])
    
    -1.09
    POSITIVE LOGITS
    .
    0.45
    ,
    0.44
    -
    0.43
     
    0.42
     (
    0.39
    ing
    0.38
    ↵
    0.37
     in
    0.36
     L
    0.36
     V
    0.36
    Activations Density 0.140%

    No Known Activations