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. 6384
    Prev
    Next
    INDEX
    Explanations

    attends from product-related tokens to the closing double square bracket token at the end of sequences

    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
    0:0.06
    1:0.08
    2:0.07
    3:0.13
    4:0.10
    5:0.05
    6:0.36
    7:0.10
    Negative Logits
    UserScript
    -0.31
    %)$
    -0.31
     Wikiseite
    -0.30
    сылкі
    -0.30
     StringTokenizer
    -0.29
     سكانية
    -0.29
    __':
    -0.29
    ISupport
    -0.29
     خارجية
    -0.28
    ագրություններ
    -0.28
    POSITIVE LOGITS
    ことで
    0.30
     препратки
    0.29
    YPE
    0.28
    Rè
    0.28
    principalColumn
    0.27
    IEU
    0.27
    aded
    0.27
    KURZBESCHREIBUNG
    0.27
    ıll
    0.27
    RegistryLite
    0.27
    Activations Density 0.004%

    No Known Activations