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-2-9B
    4. Attention Out - 16k
    5. 0-GEMMASCOPE-ATT-16K
    6. 1319
    Prev
    Next
    INDEX
    Explanations

    attends to the negation in "didn't" from a later token that represents a question or a statement expressing uncertainty or disbelief

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-att/layer_0/width_16k/average_l0_61
    Prompts (Dashboard)
    16,384 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
    0:0.03
    1:0.01
    2:0.32
    3:0.14
    4:0.08
    5:0.02
    6:0.03
    7:0.04
    8:0.01
    9:0.01
    10:0.03
    11:0.04
    12:0.02
    13:0.06
    14:0.08
    15:0.01
    Negative Logits
    '
    -0.42
    ’
    -0.28
     pertenecen
    -0.23
     parezca
    -0.23
     necesite
    -0.22
     libremente
    -0.22
     seleccione
    -0.21
     desee
    -0.21
     puedas
    -0.20
     vengan
    -0.20
    POSITIVE LOGITS
     erſt
    0.68
     zwiſchen
    0.64
     imagui
    0.64
    
    0.63
    iſchen
    0.63
     fashiola
    0.62
     EconPapers
    0.62
    linawan
    0.62
     daysTop
    0.62
    <unused52>
    0.62
    Activations Density 0.033%

    No Known Activations