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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 6463
    Prev
    Next
    INDEX
    Explanations

    attends to specific URL patterns characterized by "https://" from preceding tokens that feature the punctuation ":"

    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
    0:0.11
    1:0.14
    2:0.07
    3:0.05
    4:0.09
    5:0.35
    6:0.05
    7:0.10
    Negative Logits
    __(/*!
    -0.43
    ValueStyle
    -0.42
    TestingModule
    -0.41
    bootstrapcdn
    -0.39
     Мексичка
    -0.39
     kaynağından
    -0.38
     defaultstate
    -0.38
    weetie
    -0.37
    bewerken
    -0.37
    kuuta
    -0.37
    POSITIVE LOGITS
    ity
    0.35
    any
    0.34
    bric
    0.33
    maßen
    0.30
    ValuePair
    0.30
    ty
    0.30
     espagne
    0.29
    Alto
    0.29
    sy
    0.28
    촌
    0.28
    Activations Density 0.009%

    No Known Activations