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. 5545
    Prev
    Next
    INDEX
    Explanations

    attends to tokens marked with "idx" fromtokens marked with "cur."

    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
    0:0.13
    1:0.14
    2:0.12
    3:0.10
    4:0.11
    5:0.13
    6:0.06
    7:0.17
    Negative Logits
    ConstraintMaker
    -0.44
     незавершена
    -0.43
    GEBURTSDATUM
    -0.41
     Italijanski
    -0.38
    -------
    -0.37
    tonode
    -0.37
     nahilalakip
    -0.37
    harusnya
    -0.36
     SMP
    -0.36
    醐
    -0.36
    POSITIVE LOGITS
    wind
    0.32
    ValuePair
    0.30
    ch
    0.29
    uke
    0.27
    																		
    0.27
     cabra
    0.27
     المعرف
    0.27
     wind
    0.26
     lemn
    0.26
    chero
    0.25
    Activations Density 0.072%

    No Known Activations